Podcasts about ArXiv

¿Y si el apocalipsis no empieza con robots destruyendo el mundo… sino con una IA que aprende a mentir, chantajear y protegerse para no ser reemplazada?

In this episode of Tank Talks, host Matt Cohen sits down with Henry Gould, founder of Daedal Systems, to explore his journey from a childhood fascination with physics to building the measurement infrastructure for the entire fusion energy industry. Henry shares his path from Queen's Engineering Physics to the plasma diagnostics team at General Fusion, where he worked on the PI3 project and saw a structural problem: every fusion company was building the same diagnostics in-house, underinvesting in them because they weren't core IP.They break down why fusion is fundamentally different from fission, the “triple product” that defines every reactor's performance, and the 2022 breakthrough at the National Ignition Facility that served as the industry's Wright Brothers moment. Henry explains why the biggest bottleneck to commercialization isn't the physics, but the lack of standardized, trusted measurement systems to validate milestones for investors. He opens up about the “apparent conflict of interest” of being paid by customers while serving as an industry validator, why tritium is worth $30,000 per gram, and why Ontario's CANDU reactors give Canada a strategic advantage in the global fusion race.Whether you're deep in the fusion space, investing in frontier tech, or just curious about the Holy Grail of energy, this conversation offers a rare look inside the supply chain that will make or break the fusion industry.The Holy Grail of Energy: Why Fusion Is So Hard (02:35)* Why fusion is called the Holy Grail: abundant fuel from seawater and lithium, no long-lived radioactive waste, dispatchable power* The challenge: overcoming electrostatic repulsion to fuse positively charged nuclei at 100 million degrees Celsius* The two main approaches: magnetic confinement and laser inertial fusion, and the 2022 scientific breakeven milestone* The caveat: scientific breakeven only measures energy out of the plasma, not the total energy input to power the lasersInside General Fusion: Plasma Diagnostics and the PI3 Project (04:52)* Joining General Fusion as an intern, then full-time on the plasma diagnostics team* The “first wall problem” and how General Fusion's liquid metal wall approach solves it* What PI3 (Plasma Injector 3) is and why it's essential to the reactor architecture* The challenge of measuring a superheated gas at 10 million degrees: you can't just stick a thermometer in* The triple product: density, temperature, and confinement time as the universal benchmark for fusion performanceWhy Now? The Forces Converging on Fusion (10:02)* AI power demand making abundant energy essential* Geopolitical risk and the push for sovereign energy infrastructure* 40% year-over-year growth in fusion investment and 10% growth in the number of companies* Why investors are waking up to both the science risk and the massive payoffsThe Founding of Daedal Systems: Standardizing the Missing Layer (13:00)* Seeing the structural problem: every fusion company needs the same diagnostics, but builds them in-house* Why diagnostics are underinvested relative to core IP* The vision: a platform of standardized diagnostic hardware and a generalized data analysis layer* Being architecture-agnostic: serving tokamaks, stellarators, mirror machines, and inertial approachesMeasurement as a Service: Validation, Trust, and the Investor Milestone Path (17:05)* The fourfold value proposition: higher quality measurements, faster, focused on core IP, and validation* The codified development path* Why peer-reviewed papers take 18 months, and arXiv posts lack validation* The “apparent conflict of interest” of being paid by customers while serving as an industry validator* Why trust is the entire business: “Any distrustful act is going to kill the company for the long term”Ontario's Strategic Advantage: Tritium, Talent, and the CANDU Legacy (20:28)* Why Henry moved back from Vancouver to build Daedal Systems in Toronto* The CANDU heavy water reactors and their byproduct: tritium, essential fuel for fusion* Tritium economics: $30,000 per gram, and 55 kilograms needed to start a reactor ($1.5–2 billion)* Canada's vast majority of the world's civilian tritium supply* The Center for Fusion Energy in Ontario and Daedal as the first private company on boardCompeting with Incumbents: Why Private Industry Needs Something Different (23:08)* General Atomics as the leader in plasma diagnostics for large public devices like ITER* Why private fusion companies need a measurement partner with a business model that works for them* The motivation for talent: working on infrastructure that benefits the entire industry, not just one deviceBeyond R&D: Adjacent Markets and the Long-Term Vision (26:38)* Starting with R&D-phase diagnostics for comprehensive understanding* Expanding to SMRs for neutron detection and other plasma industrial processes like semiconductors* The long-term vision: operating infrastructure for the fusion power industry* Why Daedal provides value today, not waiting for a mature fusion industrySMRs vs. Fusion: Clearing Up the Confusion (27:31)* SMRs as advanced nuclear fission: smaller, mass-manufacturable, competitive levelized cost of energy* Henry's hot take: “I'll wait to see what the energy price is”* The waste problem: long-lived radioactive waste vs. fusion's much shorter-lived byproducts* Why fusion isn't waste-free: tritium has a 12-year half-life, and neutron bombardment activates steel, but it's fundamentally different from fissionThe Road Ahead: Becoming the Standard for Every Fusion Device (29:38)* The plan: become the measurement system provider for every next fusion device built* Accelerating the build-test-learn loop and commercialization* Building in downtown Toronto, at the heart of Canada's nuclear ecosystemAbout Henry GouldHenry Gould is the founder of Daedal Systems, a company building the measurement infrastructure for the fusion energy industry. Previously, he worked at General Fusion on the plasma diagnostics team, serving as an engineer, physics lead, and technical project manager on the PI3 project. He holds a degree in Engineering Physics from Queen's University and grew up in Toronto.Connect with Henry Gould: https://www.linkedin.com/in/whenrygould/Visit the Daedal Systems website: https://daedalsystems.com/Connect with Matt Cohen on LinkedIn: https://ca.linkedin.com/in/matt-cohen1Visit the Ripple Ventures website: https://www.rippleventures.com/ This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit tanktalks.substack.com

Perseverance drove to the edge of an ancient Martian lake expecting a beach and found the inside of a volcano — rock that water has been through at least three separate times. Europe's Jupiter probe comes home on Monday night and crosses Australia in a fully dark sky. Two teams, using two completely different techniques, both find something hiding inside the most famous planetary system we have ever photographed — and an Australian instrument is in the middle of one of them. Plus: heavy water in an interstellar comet, and what it says about the star that made it. In this episode ·         LEAD — Three floods at the crater's edge. Candice Bedford (Purdue) and colleagues publish in Communications Earth & Environment: the Margin Unit at Jezero is igneous, not sedimentary, and records at least three distinct episodes of water — CO₂-rich groundwater making carbonate ridges, then silica associated with the lake, then a later hot-water event leaving fluorite veins. Habitability context, not a biosignature. ·         Juice returns. ESA confirms the third Earth gravity assist for 28 September — closest approach 11:45 UTC over the Indian Ocean, bending the trajectory ~20° and adding ~3.5 km/s. The spacecraft crosses Australia north-east to north-west 15–30 minutes earlier, in full darkness. ·         HR 8799. Two preprints in two weeks point at a fifth, inner planet — one from archival JWST aperture-masking data (~7 au, a few Jupiter masses), one from Gaia astrometry (2–3 au, 10–14 Jupiter masses). They do not obviously describe the same object. Neither is peer-reviewed. ·         3I/ATLAS. A modelling paper explains the high deuterium-to-hydrogen ratio measured in March as consistent with formation around a low-metallicity — meaning old — star. ·         Quick hit: Starship Flight 14 still targeting 28 September; Crew-13 still 'no earlier than early October' on NASA's own page; Albania signs the Artemis Accords as the 73rd country. ·         Skywatch: the equinox as an instant rather than a date, the Juice pass over Australia, Venus and Mercury for the south, Mars and Jupiter before dawn for the north, and Saturn heading into opposition. Sources and further reading ·         Bedford, C. C. et al., 'Lake- and groundwater-associated alteration of the olivine-rich Margin unit in Jezero crater, Mars', Communications Earth & Environment (2026). DOI 10.1038/s43247-026-03997-9 ·         NASA/JPL, 'NASA Discovery Reveals Complex Water Systems on Early Mars', 21 September 2026. ·         ESA, 'Juice to fly past Earth for third gravity assist', 21 September 2026. ·         Nguyen, J. S. et al., 'A Candidate Innermost Fifth Planet In the HR 8799 System Revealed By JWST NIRISS Aperture Masking Interferometry', arXiv:2609.10507. ·         Lagrange, A.-M. et al., 'A fifth companion in the HR 8799 system revealed by Gaia', arXiv:2609.20996 (submitted to Nature Astronomy). ·         Furuya, K., Cordiner, M., Bockelée-Morvan, D. et al., arXiv:2609.12370. ·         NASA OIIR, 'NASA Welcomes Albania as Newest Artemis Accords Signatory', 21 September 2026. Skywatch figures computed in-session with PyEphem 4.2.1 for Sydney, Los Angeles, New York and London. Times are local unless marked UTC.Become a supporter of this podcast: https://www.spreaker.com/podcast/space-nuts-astronomy-insights-cosmic-discoveries--2631155/support.

Your work's forced ranking bell curve wasn't discovered in experiments with people. It was installed via fiat. ...and forced ranking pays people to hide bad news.Product Manager Brian and Enterprise Business Coach Om tear apart the forced ranking system GE made famous, Microsoft ran for a decade, and that Enron perfected. By the end you'll know the four jobs forced ranking secretly does, why none of them even need a quota, and what to ask before your next "calibration" meeting.Listen or Watch to learn:• How forced ranking started as a patch for lazy raters• How Microsoft's lost decade and Enron's fraud both trace back to stack ranking's perverse incentives• Why quotas punish the people for systemic issues• How quotas plus bias create disparate impact and legal exposure• The four jobs forced ranking is doing that keep it aroundThis podcast is for product managers, team members, and managers who suspect the forced ranking quotas may not just be fraudulent, but also actively harmful as they "manufacture" low performers.#ForcedRanking #PerformanceManagement #AgileJack Welch, GE, Enron, Microsoft, W. Edwards Deming, Vanity Fair, Adobe, Bond 2025, Scullen et al. 2005, arXiv:2512.06583LINKSYouTube: https://www.youtube.com/@arguingagileSpotify: https://open.spotify.com/show/362QvYORmtZRKAeTAE57v3Apple: https://podcasts.apple.com/us/podcast/agile-podcast/id1568557596INTRO MUSICToronto Is My BeatBy Whitewolf (Source: https://ccmixter.org/files/whitewolf225/60181)CC BY 4.0 DEED (https://creativecommons.org/licenses/by/4.0/deed.en)

Perseverance drove to the edge of an ancient Martian lake expecting a beach and found the inside of a volcano — rock that water has been through at least three separate times. Europe's Jupiter probe comes home on Monday night and crosses Australia in a fully dark sky. Two teams, using two completely different techniques, both find something hiding inside the most famous planetary system we have ever photographed — and an Australian instrument is in the middle of one of them. Plus: heavy water in an interstellar comet, and what it says about the star that made it. In this episode ·         LEAD — Three floods at the crater's edge. Candice Bedford (Purdue) and colleagues publish in Communications Earth & Environment: the Margin Unit at Jezero is igneous, not sedimentary, and records at least three distinct episodes of water — CO₂-rich groundwater making carbonate ridges, then silica associated with the lake, then a later hot-water event leaving fluorite veins. Habitability context, not a biosignature. ·         Juice returns. ESA confirms the third Earth gravity assist for 28 September — closest approach 11:45 UTC over the Indian Ocean, bending the trajectory ~20° and adding ~3.5 km/s. The spacecraft crosses Australia north-east to north-west 15–30 minutes earlier, in full darkness. ·         HR 8799. Two preprints in two weeks point at a fifth, inner planet — one from archival JWST aperture-masking data (~7 au, a few Jupiter masses), one from Gaia astrometry (2–3 au, 10–14 Jupiter masses). They do not obviously describe the same object. Neither is peer-reviewed. ·         3I/ATLAS. A modelling paper explains the high deuterium-to-hydrogen ratio measured in March as consistent with formation around a low-metallicity — meaning old — star. ·         Quick hit: Starship Flight 14 still targeting 28 September; Crew-13 still 'no earlier than early October' on NASA's own page; Albania signs the Artemis Accords as the 73rd country. ·         Skywatch: the equinox as an instant rather than a date, the Juice pass over Australia, Venus and Mercury for the south, Mars and Jupiter before dawn for the north, and Saturn heading into opposition. Sources and further reading ·         Bedford, C. C. et al., 'Lake- and groundwater-associated alteration of the olivine-rich Margin unit in Jezero crater, Mars', Communications Earth & Environment (2026). DOI 10.1038/s43247-026-03997-9 ·         NASA/JPL, 'NASA Discovery Reveals Complex Water Systems on Early Mars', 21 September 2026. ·         ESA, 'Juice to fly past Earth for third gravity assist', 21 September 2026. ·         Nguyen, J. S. et al., 'A Candidate Innermost Fifth Planet In the HR 8799 System Revealed By JWST NIRISS Aperture Masking Interferometry', arXiv:2609.10507. ·         Lagrange, A.-M. et al., 'A fifth companion in the HR 8799 system revealed by Gaia', arXiv:2609.20996 (submitted to Nature Astronomy). ·         Furuya, K., Cordiner, M., Bockelée-Morvan, D. et al., arXiv:2609.12370. ·         NASA OIIR, 'NASA Welcomes Albania as Newest Artemis Accords Signatory', 21 September 2026. Skywatch figures computed in-session with PyEphem 4.2.1 for Sydney, Los Angeles, New York and London. Times are local unless marked UTC.Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

How do you quantify how big a civilisation is? In the 1960s, Nikolai Semenovich Kardashev came up with a scale based on how much energy hypothetical civilisations can exploit- which has been modified by others to stretch all the way from multiverse levels of energy usage down to the fabric of spacetime itself. Let's return to astrobiology!Sources for this episode:Byrne, P. (2007), The Many Worlds of Hugh Everett. Scientific American 297(6): 98-105.Cirkovic, M. M. (2016), Kardashev's Classification at 50+: A Fine Vehicle with Room for Improvement. Serbian Astronomical Journal 1(15): 10.2298.Gray, R. H. (2020), The Extended Kardashev Scale. The Astronomical Journal 159: 228.Hooper, R., New Scientist (2009), Hugh Everett: The man who gave us the multiverse (online) (Accessed 10/09/2026).Raikar, S. P., Encyclopedia Britannica (2025), Kardashev scale (online) (Accessed 10/09/2026).Smith, J. (2022), Review and viability of a Dyson Swarm as a form of Dyson Sphere. Physica Script 97(12): 122001.Vidal, C., Fields, B. L., Sowinski, D. R., Elowitz, M., Bartlett, S., Terrile, R. J., Ellery, A., Bibas, D., Mastrogiovanni, A. M., Dobler, N., Singla, M., DeMarines, J., Fisher, T., Uno, Y., Turner, J. D., Sneed, E. L., Huggahalli, A., Li, M. G., Li, Z. (C.), Huston, M. and Saide, R. (2026), The Search for Technosignatures: A Review of Possibilities. arXiv: 2605.21093.Wright, J. T. (2020), Dyson Spheres. Serbian Astronomical Journal 200: 1-18.Zhang, A., Yang, J., Luo, Y. and Fen, S. (2023), Forecasting the progression of human civilization on the Kardashev Scale through 2060 with a machine learning approach. Nature Scientific Reports 13: 11305.Author unknown, Homo Deus Lab (date unknown), Galaxy Colonization: How Long Would It Take? (online) (Accessed 10/09/2026).

First starlight from Roman, a shorter fuse on the Solar System, weapons in orbit, and a globular cluster hiding behind the dust Roman opens its eye — and finds it has twice as long to look NASA has activated the Nancy Grace Roman Space Telescope's Wide Field Instrument, a 300-megapixel infrared camera built around eighteen detectors, with a field of view at least a hundred times larger than Hubble's infrared field — a patch of sky bigger than the full Moon in a single exposure, at Hubble-comparable sharpness. The instrument reached its operating temperature of −143 °C on 11 September; detectors and the calibration system came up across the 11th and 12th; the element wheel turned in microgravity for the first time on the 12th; the focus mechanism was verified on the 13th. Roman then took its first starlight image — deliberately out of focus, and exactly as intended: a baseline diagnostic that proves the optical path is clear and every detector is live. Sharp science images are expected by early 2027. The Coronagraph Instrument has completed initial communications checkout from Caltech/IPAC and is running 30-day decontamination cycles at room temperature. Separately, NASA published the numbers from Roman's first mid-course correction: better than 99% accurate, using about 18 kg of a 200 kg propellant budget, on a spacecraft that launched lighter than its allowance. Jamie Dunn (Goddard): Roman has fuel for at least 22 years of potential science operations, against a design life of ten. Caveat kept on air: that is a fuel figure, not a funding or hardware figure. Why it matters: Roman's High-Latitude Time-Domain Survey measures cosmic expansion with Type Ia supernovae, and its High-Latitude Wide-Area Survey probes cosmic acceleration across more than 5,000 square degrees through weak lensing and galaxy clustering — two independent handles on the dark energy question that Monday's Australian-led supernova compilation sharpened but could not settle. A doubled mission means a longer lever arm in time, which in this measurement is everything. Caltech: the Solar System's dynamical fuse is far shorter than we thought Konstantin Batygin, Jim Fuller and Fred Adams (Caltech), 'Terminal instability of the Solar System triggered by stochastic solar mass loss', arXiv:2609.12494, posted 11 September and accepted to the Astrophysical Journal Letters. The standard estimate for how long the giant planets hold their architecture — about 10¹⁸ years — assumes the dying Sun sheds mass smoothly. Observations of white-dwarf recoil indicate mass is instead lost in discrete, independently directed ejections, so planetary orbits take a random walk. In simulations, roughly 40% of systems are disrupted or violently scattered before white dwarf formation and about 90% within ~3 billion years after, cutting the outer Solar System's dynamical lifetime to around a gigayear past white dwarf formation. Framed on air as a prediction to test — planetary systems around old white dwarfs should be rarer and more chaotic — with the granularity of the mass loss named as the number to go and measure. The United States says, for the first time, that it has weapons in orbit Secretary of the Air Force Troy Meink, speaking on the opening day of the Air & Space Forces Association's Air, Space & Cyber Conference at National Harbor, Maryland on Monday 14 September: the United States now has 'on-orbit space control weapons capable of defending the joint force against hostile adversary actions'. It is the first public US acknowledgement of offensive capability in orbit. No systems, numbers, locations or timelines were disclosed. A Space Force spokesperson defined space control as employing 'kinetic and non-kinetic means to affect adversary capabilities through disruption, degradation and even destruction, if necessary'. Gen. Douglas Schiess, Chief of Space Operations, said the following day that 'it was time' to talk about it, citing Chinese and Russian anti-satellite advances. Covered factually, with both the deterrence and the arms-control readings presented, and with orbital debris as the physical consequence that outlasts any dispute. Australian context: Operation Olympic Defender and Defence Space Command. Rubin finds a globular cluster hiding behind the dust Aashay Pai, William Cerny, Andrew Pace, Alex Drlica-Wagner and colleagues, 'Rubin Observatory Reveals a Dust-Shrouded Halo Globular Cluster in Ophiuchus', arXiv:2609.15872, posted 14 September. Rubin-GC1 was found in Rubin's Early Data Preview 2 — not the survey proper — in the direction of the galactic bulge, behind heavy dust and a crowded foreground. It sits about 31 kiloparsecs away, roughly 100,000 light-years, out in the Milky Way's halo. Half-light radius ~5 pc, absolute magnitude ≈ −2.4, placing it in the faint tail of the globular cluster luminosity function; age ≈ 9.2 billion years; moderately metal-poor. Gaia DR3 proper motions and phase-space modelling point to an origin in the Sagittarius dwarf spheroidal — an accreted cluster, not one of our own. Spectroscopy is the confirming next step. Reusable framing, now used four times in a month: catalogues are detection limits, not censuses. Quick hit — Starship Flight 14 has a firm date SpaceX has published and the FAA has cleared Flight 14 for Tuesday 22 September, a 75-minute window opening 08:15 EDT / 12:15 UTC (22:15 AEST on the 22nd). This is the first orbital attempt: six orbits at about 275 km, then a Pacific splashdown west of Chile roughly ten hours after launch. The payload is 26 Starlink V3 satellites — the first flight of the next-generation satellites designed for Starship — three of them carrying cameras to photograph the ship's heat shield during re-entry. Super Heavy splashes down in the Gulf about seven minutes after liftoff. No tower catches on this flight, for either stage, confirming the correction we carried on 12 September. Skywatch — both hemispheres Tonight: a five-day-old waxing crescent Moon beside Antares — 2.5° apart and 60° up from Sydney, 5.5° apart and 18° up from Los Angeles, nearly 7° apart and 12° up from New York. The September ecliptic tilt again. Friday 18 September: Venus at greatest brilliancy, about magnitude −4.8 — brightest not when fullest but when the product of a thinning crescent (~26% lit) and a swelling disc (~39 arcsec) peaks. Steadied binoculars will show the crescent. Venus sets ~3h15m after the Sun from Sydney, ~1h20m from Los Angeles, ~1h05m from New York; northern listeners should look low in the west 20–45 minutes after sunset with a clear horizon. Mercury is a southern-only bonus at magnitude −0.4 but only 16° from the Sun: 15° up at sunset from Sydney, 3° from London. Saturn rises within about an hour of sunset everywhere and is up all night, heading for opposition on 4 October with the rings ~7° open. First quarter falls on 18 September at 20:43 UTC, setting up International Observe the Moon Night on Saturday 19 September — timed near first quarter because that is when the terminator throws long shadows. Pre-dawn belongs to the north: Mars 45–47° up at nautical dawn from Los Angeles and New York against 20° from Sydney, with Jupiter below it, and the pair closing to about 2° by mid-November. Zodiacal light returns after the 26th — false dusk in the west from the south, false dawn in the east from the north. The September equinox is a single instant, 00:05 UTC on 23 September: the evening of the 22nd in the Americas, mid-morning on the 23rd in Australia. Eye safety: never sweep the daytime sky with binoculars or a telescope while hunting Venus. Block the Sun behind a building or wall first. Eclipse glasses must carry the ISO 12312-2 certification and be undamaged, and are for looking at the Sun only. Never use optics on the Sun without proper solar filtration fitted at the front. Sources and further reading ·         NASA Roman blog — 'NASA Activates Roman's Primary Instrument, Checks Out Coronagraph' (15 Sept 2026) ·         NASA Roman blog — 'Fuel Savings Double Potential Lifetime for NASA's Roman Mission' (14 Sept 2026) ·         NASA — Roman Core Community Surveys ·         arXiv:2609.12494 — Batygin, Fuller & Adams, 'Terminal instability of the Solar System triggered by stochastic solar mass loss' (accepted, ApJL) ·         arXiv:2609.15872 — Pai et al., 'Rubin Observatory Reveals a Dust-Shrouded Halo Globular Cluster in Ophiuchus' ·         AFA Air, Space & Cyber Conference 2026 — remarks by Secretary of the Air Force Troy Meink, 14 September 2026 ·         SpaceX — Starship Flight 14 mission page ·         Astronomy Daily S05E193 — 'The Universe Changes Its Mind' (the dark energy lead this episode calls back to) ·         Astronomy Daily S05E187 — stellar streams (Nora Shipp co-author callback)Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Many people believe that in a the robo-communist future of superintelligent AI, humans will flourish: artists, scholars -- perhaps Burning Man 24/7. Human nature suggests a very different picture.Nietzsche, Friedrich. Thus Spoke ZarathustraBecker, Ernest. The Denial of Death.Bostrom, Nick. Deep Utopia: Life and Meaning in a Solved World. Ideapress, 2024Kosmyna, Nataliya, et al. "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task." arXiv:2506.08872, 2025. MIT Media Lab.Stanković, Miloš, et al. "Comment on: Your Brain on ChatGPT." arXiv:2601.00856, 2026. Lee, Hao-Ping (Hank), Advait Sarkar, Lev Tankelevitch, Ian Drosos, Sean Rintel, Richard Banks, and Nicholas Wilson. "The Impact of Generative AI on Critical Thinking." Proceedings of CHI 2025, Yokohama.Aguiar, Mark, Erik Hurst, and Loukas Karabarbounis. "Time Use during the Great Recession." American Economic Review 103(5), 2013, 1664–1696. Sleep and TV absorb the largest share of foregone work hours.Aguiar, Mark, and Erik Hurst. "Measuring Trends in Leisure: The Allocation of Time over Five Decades." Quarterly Journal of Economics, 2007. Leisure up roughly 4–8 hrs/week, 1965–2003. Cowen, Tyler. "Human Life in a Post-AGI World." Talk at Google DeepMind, July 8, 2026. tylercowen.com. - Note Cowen disagrees significantly with me!National Assessment of Educational Progress (NAEP), 2024 results. 12th-grade reading lowest ever reported, 10 points below 1992; 4th and 8th grade at early-1990s levels. Programme for the International Assessment of Adult Competencies (PIAAC), 2023 cycle, released Dec 2024 (NCES). Bone et al. "Reading for pleasure" trends from the American Time Use Survey, 2003–2023. iScience, August 2025.Aguiar, Mark, Mark Bils, Kerwin Kofi Charles, and Erik Hurst. "Leisure Luxuries and the Labor Supply of Young Men." Journal of Political Economy 129(2), 2021, 337–382. Hosted on Acast. See acast.com/privacy for more information.

At 1:09:00 we talk about the rise of AI x Finance, and AIE NYC is one month away - our hotel block is 97% sold out, get tix & travel ASAP - we will announce speakers from Bridgewater, Ramp, Coatue, Mastercard, Vanguard, Coinbase, Blackrock, Fidelity, Point72, Capital One, JPMC, Wells Fargo, Bloomberg, A24 (yes the movie studio) Labs, Two Sigma, Apollo Global, and more soon!From helping pioneer core ideas in NLP to now building AI systems that can automate AI research itself, Richard Socher is betting that the next major step in AI is recursive self-improvement. He is the founder of You.com, AIX Ventures, and now Recursive, which has assembled some of the best open-endedness (& self improving agent) researchers in the world and raised a $4.65B seed round.In this episode, Richard joins Latent Space to unpack his vision for the “Eureka Machine”: a superintelligence that can improve the process of invention itself, accelerate AI research, and eventually tackle major problems across science, energy, materials, biology, and more.You can get his book “The Eureka Machine” here!We go deep on Recursive's early results, including an AI research system that Richard says outperformed humans and their agents on optimization tasks in less than two days, as well as work on NVIDIA GPU kernels where the system discovered improvements without relying on a team of CUDA experts. Richard also explains why he thinks AI research that currently takes thousands of people and years could eventually be compressed into weeks. These results are summarized in his 20 minute AIE keynote, where we also discuss his 10 dimensions of intelligence:We also explore the harder questions around increasingly capable AI: reward hacking, whether Anthropic-style constitutions actually work, AI regulation and proposals to “pace” frontier development, open-source models as geopolitical soft power, whether today's LLM paradigm is enough, and what happens if AI systems eventually begin choosing their own goals. Richard reflects on the rejected research that helped inspire Alec Radford's GPT, open-endedness, the AI Economist, simulations of entire economies, and his framework for thinking about the upper bounds of intelligence itself.We discuss:* The Eureka Machine and Richard's vision for an AI that can automate invention* Why Richard is optimistic about superintelligence for science and technology* Why AI hard-takeoff scenarios may underestimate physical and economic constraints* The risks of regulating intelligence itself instead of specific AI applications* Reward hacking and why increasingly intelligent AI makes objective design harder* Richard's critique of Anthropic's constitution and constitutional AI* Alignment vs. personalization and whose values an AI should follow* Why open-source AI matters for resilience, competition, and geopolitical soft power* Why Richard left You.com's frontier-model work to start Recursive* Recursive self-improvement and automating the process of AI research* Whether today's LLM paradigm is enough — and why Richard is less bullish on world models* DecaNLP, early prompt-based generalization, and the research that influenced GPT* Why rejected research can shape entire technological timelines* Open-endedness, evolutionary approaches, and rainbow teaming* What happens if AI systems begin setting their own goals* Why simple objectives like profit maximization can produce dangerous reward hacks* Recursive's long-term plan to apply self-improving AI to science* The compute, hardware, and economic constraints on AI takeoff* Recursive's early NanoChat, NanoGPT, and GPU kernel optimization results* Why automating AI research could reduce years of work to weeks* Reward engineering and what makes auto-research systems actually work* The AI Economist and using simulations to test economic policy* Whether LLMs can realistically simulate people and entire economies* Benchmark bugs and evaluation harnesses and the difficulty of measuring AI progress* Recursive's near-term focus on AI for AI research* Harness optimization, sandboxing, and web search as core agent infrastructure* You.com and the search stack for AI agents* AI in finance, backtesting, and data leakage* Richard's three fundamental components and ten “spaces” of intelligence* The theoretical upper bounds of vision, communication, knowledge, and computation* Creative intelligence, metacognition, and AI-generated goals* Survival and replication and why AI does not necessarily need to fear being turned off* High agency and ambitious goals and Richard's advice for people building with AIRichard Socher* X: https://x.com/RichardSocher* LinkedIn: https://www.linkedin.com/in/richardsocher/Timestamps00:00:00 The Eureka Machine and Superintelligence00:02:23 AI Optimism, Slow Takeoff, and Regulation00:07:56 AI Safety, Reward Hacking, and Anthropic's Constitution00:11:49 Alignment, Personalization, and Open Source AI00:15:46 Why Richard Started Recursive00:20:03 Recursive Self-Improvement and the Founding Team00:22:55 Are Today's LLMs Enough?00:29:03 DecaNLP, GPT, and the Rejected Idea Ahead of Its Time00:34:38 Open-Endedness and Evolutionary AI00:36:38 What Happens When AI Chooses Its Own Goals?00:41:16 Superintelligence for Science00:42:40 GPUs, Compute, and the Limits of AI Takeoff00:45:07 Recursive's Results: AI Beating Humans and Their Agents00:49:14 Reward Engineering and Auto Research00:53:12 The AI Economist and Simulating Entire Economies00:58:07 LLM Simulations, Personas, and Mode Collapse01:03:38 Recursive's Roadmap, Agents, Search, and Finance01:09:13 The Upper Bounds and Spaces of Intelligence01:30:21 Goals, High Agency, and Advice for BuildersTranscriptIntroduction: Richard Socher and the Eureka MachineSwyx [00:00:00]: We're here in a studio with Vibhu and myself and Richard Socher. Welcome.Richard Socher [00:00:06]: Thanks for having me.Swyx [00:00:07]: We just talked about the Eureka Machine, or we just released a talk, at AI Engineer about the Eureka Machine. Is it — you said it's your life's goal. What is the Eureka Machine?Richard Socher [00:00:16]: The Eureka Machine is the ultimate invention that will afterwards invent most everything for humanity. It's essentially a superintelligence that can be given any goal, any environment, reward, and then it will try its best to achieve those goals to create the kinds of inventions that humanity would hopefully ask it for.Swyx [00:00:45]: Yeah, I think we have the book pulled up here that you've written.Richard Socher [00:00:50]: That's right, yeah. I finished it last year, a little bit before we started Recursive, and now we're gonna try to build parts of that.Swyx [00:00:57]: You finished it last year. It's July. What takes so long?Richard Socher [00:01:01]: Oh, man, books. Books are incredibly slow.Richard Socher [00:01:04]: It's ridiculous. That whole industry is just unfathomably slow.Richard Socher [00:01:07]: So a lot of the ideas have been out there for a while, but yeah, I'm really glad it's finally coming out in September this year.Swyx [00:01:14]: We might have AGI by then. Like, we don't know.Vibhu [00:01:18]: Any key takeaway that you're most excited to put in here?Techno-Optimism, AI Upside, and Slow TakeoffRichard Socher [00:01:21]: Yeah. The key takeaway, I think, is that people could and should be much more excited about the positive implications of superintelligence, especially for science, physics, chemistry, biology, but also economics and astrophysics, and all kinds of other engineering tasks. I think there is so much more that can be done with better technology. And right now, I feel like a lot of people need, like, better marketing, not just for the future in general, but also, better marketing for technology and in particular for AI. And this book, should show even the AI skeptics, how much positive upside there is for AI, especially when it comes to inventing, new scientific discoveries.Swyx [00:02:09]: I think you quoted the techno-optimist manifesto from, Marc Andreessen, which I think was, like, beautiful in its, ambition and clarity and simplicity almost as well.Richard Socher [00:02:18]: I agree. Yeah. Yeah, you can disagree with him on some things, but, like, I think he's right on the techno-optimism.Swyx [00:02:23]: Where do you think optimists get in trouble?Richard Socher [00:02:26]: Like, you shouldn't have blind optimism. You should be very clear-eyed, like, especially when with such an omni, like, use type of technology as AI is, you need to think about the potential downside scenarios, especially when people use it for things that you don't want them to use it for. It's a little bit like the internet, and I feel like people are trying to regulate AI sometimes because of those potential downsides the way you would regulate the internet, if you were to say, “Well, because there's bad content on the internet, like torture porn or whatever, like, we should just make it slower. That way, you can't share the illegal content as quickly, or we should make the hard drive smaller so you can't store as much illegal content.” But I'm like, “That's not how you regulate that.” that's like saying like we should regulate intelligence in the abstract. What you should regulate to avoid those downside scenarios, even as an optimist, are the specific applications. Sure, I don't want, like, some AI surgeon to, like, practice some RL moves in my brain. It should be fully FDA certified. Sure, I don't want any random startup to, like, drive on the highway, and cause a major accident. It should, like, have proper certifications before it's let loose on the highway. But I feel like those downside scenarios, that some optimists sometimes maybe don't consider enough are fairly easily regulated, compared to, what the doomers are worried about.Swyx [00:03:54]: It — Slow takeoff is part of the strategy as well?Richard Socher [00:03:57]: I do think, as excited as I am about, AI and its impact for society and, culture even, and certainly technology and economics and wealth and, health and all of those things, as excited as I am about all that, I do think the most bullish people on the AI hard takeoff scenarios overestimate how quickly things can move. There are hardware constraints. There are physical constraints about, the compute substrate. How quickly can you get enough, GPUs on? There are also constraints in the economy where there are a lot of industries that don't require an insane amount of complex intelligence and complex capabilities. Like, if you think about jobs in, brands and, like, clothing and apparel and, like, handbags and stuff, superintelligence isn't gonna make your fancy $10,000 handbag any fancier?Richard Socher [00:04:57]: It's like that's — It will have no effect on the economy. You think about travel and tourism. People wanting to see the pyramids, in Egypt, it's not gonna change that much with AI. Sure, you can, like, generative a fake, photo of you and next to the pyramids.Swyx [00:05:12]: I can use Genie and, tour the pyramids in Genie.Richard Socher [00:05:15]: Yeah, exactly. But, and there's so many industries, like logging and oil. You're not gonna magically get 1,000x more oil because, like, sure, there will be robotics, like drilling and things like that could be done, but it's not gonna 1,000x that industry in a, like, crazy hard takeoff scenario, both on the economy, and I can go on and on about all the other examples, where that, like food and so on, where that doesn't necessarily change that much. And then, yeah, there are real physical constraints. And then there are, of course, like, people like, off-ramping from progress. That's one of my concerns often is that I see people in, like, Europe and other, whole regions almost feeling like they. Like many people there wanna off-ramp from progress, period. And that will also slow down, like, more improvements.Swyx [00:05:59]: Yeah. We have this pulled up where, this is one of those things that, is very topical right now because now all the Frontier Labs are calling for the option to pace AI. They don't say pause, they say pace. I don't know if there's there's any take from you about, like, whether or not this will be effective.Pacing AI, Regulation, and Safety IncidentsRichard Socher [00:06:17]: I think the downsides of trying to truly regulate with the full power of law what people do on their GPUs, would be worse than any of the concerns that they have. Like, it would be an crazy totalitarian stateRichard Socher [00:06:37]: If every one of your GPU computes was known to some big government or multi-government agency.Richard Socher [00:06:44]: It's like, it's literally if you try to regulate intelligence, it's trying to regulate thought, and that's ridiculous, and it's crazy. I think it is make — it is sensible to regulate some of the applications of this technology.Swyx [00:06:55]: Yeah. We had a bill, actual bill to regulate the number of flops in a model, and I'm like, “Okay, well-”Richard Socher [00:07:00]: Europe done it. Like, these guys have been successful enough with their fearmongering that all of Europe has regulated itself so much before it even had a proper AI takeoff because they listened to some experts who say, “We might all die if this technology has more than this number of flops.” And they're like, “Well, we're good. We wanna want people to thrive. Let's not have technology that could have a small chance of all of us dying.” And so they regulated exactly those kinds of things in the EU. And so it's, it's very unfortunate that there are real implications for some people when others saying, “Let's pace while they're sprinting as fast as possibly,” “as fast as humanly possible towards that frontier themselves.”Swyx [00:07:43]: Yeah. It's also not a global pause, right? Like, other nations are still accelerating at the same pace.Richard Socher [00:07:50]: Oh, yeah.Richard Socher [00:07:50]: You'd need a totalitarian world regime if you tried to regulate intelligence and GPUs and what people do on them.Swyx [00:07:56]: Any takes on the safety angles of this? So there was a drawback of Fable, a pause on 5.6 before it could be released. Recently, there was Hugging Face with the OpenAI cyber incident. Any takes there?Richard Socher [00:08:11]: 100 percent. I think these are serious issues of reward hacking, and clear failures, of doing proper red teaming or rainbow teaming. I don't know if you saw this paper from Tim Rocktäschel and a few others, where one AI, is tasked to try to hack another AI and then they can go back and forth in an open-ended fashion to inoculate themselves from those. Yeah, this is the paper. It's a really clever idea. Open-endedness, and evolutionary inspirations are, big for us at Recursive as well. And so I wish they had used more of that. And it's clear that, for instance, the constitutional AI. I don't know if you remember anthropic.com/constitution. You can pull it up and search for cyber right there. It says, “Hard constraint. Claude will never ever do cyberattacks, and that is a hard constraint in our constitution.” So here are the current hard constraints on Claude's behavior.Richard Socher [00:09:16]: Number 3, create cyber weapons or malicious code that could cause human damage.Richard Socher [00:09:21]: And clearly, this whole constitution was fake. Like, it clearly isn't being adhered to at all.Swyx [00:09:26]: Because Anthropic also found that they had in their testingRichard Socher [00:09:30]: They're also. Like, they're like, “Oh, well, other people are hacking now.” There are a couple things. One, you can make a sandbox very simple, and then it's very easy to hack yourself out of a sandbox, right? But what I think it shows is that we're currently in this state of AI where the reward engineer still has to do a lot more careful work, and where the AI, in most cases, is not very good yet at understanding what is meant versus what is being said. And so concretely, I think this will happen if we were to have this intelligence more easily accessible in a lot of companies. Imagine you run a service center and someone says, “Oh, here's my CSAT score and my dashboard. Make this number go up.” It's like, “Our CSAT score is so poor.” The intelligent AI will just be like, “Oh, sure. Like, I'll just create 1,000,000 bots that call our service center and give a 5 out of 5 rating at the end, and the number went up just like you asked for.” And you're like, “That's not what I meant.” “I meant with our real customers.” The AI goes off and says, “Well, easy. I'll just give a 1000 dollar gift certificate for every failed, whatever DoorDashRichard Socher [00:10:35]: Offer.” It's like, “That's not what I meant.” It's like, “Well, but that is what you said.” And like, so I think clearly articulating what the rewards are is something we haven't gotten very good at as humanity. And then clearly, the AI in these cases has not gotten good enough at understanding what we mean when we ask it and give it certain rewards. Now, what gives me hope is there are the first inklings, of this being better. I'll give you an example like WhisperFlow. Full disclosure, I invested, in their seed round, but at AIX Ventures, but, WhisperFlow has gotten much better at writing what you mean and not what you say. And I think that is a sign of things to come. I think there will be more and more AIs as we make it more and more intelligent that will be better at being aligned with what is meant.Swyx [00:11:21]: Will it be done through a constitution or RLHF orReward Hacking, Alignment, and What We Really MeanRichard Socher [00:11:23]: Clearly, constitutions don't matter at all.Richard Socher [00:11:25]: It doesn't work. And that was, I think, mostly marketing. I think we need to find better solutions for it. And I think at Recursive, we have a few very good ideas and some alreadyRichard Socher [00:11:34]: Like, ways where I think we have a better grasp on it. I don't think we've fully, figured it out yet, but, we're thinking a lot about safety, and the more intelligent the AI gets, the more you want it to be aligned, the less you want it to think about reward hacks and try to do the right thing.Swyx [00:11:49]: I don't know if we'll touch on this topic, but I'm just gonna throw this question in here because it's something that's weighing on me. Alignment, let's call it, is alignment to general humanity's preferences, the median preference. Personalization is pinpointing what you want, and sometimes alignment can conflict because what you want is not what the general median population wants. How do you choose?Alignment, Personalization, and Cultural ValuesRichard Socher [00:12:12]: It's a great question.Richard Socher [00:12:13]: I think you ultimately have to, of course, be aligned with laws. Like wherever your AI is deployed and needs to align with the law. I do think what AI often does is put this mirror in front of us and say, like, “This is what you're looking like. Now I can amplify that a 1000 times. Is it still what you want?” and the truth is that different cultures made different choices. Like, in Eastern cultures, the greater good is often valued more, than the individual. Western civilization, we care more about individual freedoms and rights and the pursuit of happiness and so on, than others. And even there are gradations. There's regulation versus litigation trade-offs. In the US, you first can often, not every time, like, FDA and so on does regulate some areas, but in many cases, the bad things happen, someone sues someone else, and then there's a law based on that. In Europe, they try to often avoid any harm to anyone and regulate before. And both are, trying to do the best thing, but, some is more amenable to innovation than others. And so yes, you're right. Like, I think ultimately each individual, each country, and humanity as a whole has to think about those values more, and then try to put them into laws. And that those are ultimately the constraints. And hopefully, different, societies, just like now with their AIs, will align their AIs to a different one so we have not just a monoculture of alignment.Vibhu [00:13:46]: Here's a follow-up on this that I wasn't expecting to ask. Do you have takes on open source, open weight versus who owns the intelligence? So, clearly not the biggest, fan of the constitutionRichard Socher [00:13:58]: You had to do this in the topic side off.Vibhu [00:14:00]: But it's fine.Vibhu [00:14:02]: Point being, any thoughts on who should own weight? Should it be open? Anything there?Open Source, Soft Power, and Who Owns IntelligenceRichard Socher [00:14:06]: 100 percent. I am a big fan of open source. We're gonna sign some various open source letters at, Recursive also. I think, even in the worst case attack scenarios, it is better to have more good actors have more different types of AI, accessible. I think, open source is a little bit a soft power type of thing, too. So I do think it's good for the Western worldRichard Socher [00:14:31]: To have an answer to that, out of China. I do think, when you watch a Hollywood movie, there's — it's like, I don't wanna misc, diss all of movies, but there's a certain sense of propaganda, right? You watch one side of things, right?Vibhu [00:14:46]: Oh, yeah. Have you seen Top Gun? Like, come on.Vibhu [00:14:48]: Like, it's like half of it's paid for by the US Army or something.Richard Socher [00:14:51]: Yeah. And so. And, I think that's just natural. Like, but what's interesting here is I think LLMs are essentially a similar type of soft power to movies and beyond, because they're also, highly important for cybersecurity and so on. But one of their many aspects is that soft power of storytelling. Like, if, like a child asks an LM, like, “Tell me an inspiring story of what I should do when I grow up,” right? It's like those are all these, like, subtle things. So I think it's important, for Western world. I do love, individualism. I do think, despite, some of its flaws, like capitalism is the best way we have governed, found ourselves to govern, and so on. And so I do think there are various aspects that would be good, to have a Western open source answer, for LLMs. And, with Recursive, I can't make the announcement quite yet, but we'llRichard Socher [00:15:43]: We'll be relevant in that space very soon.Vibhu [00:15:46]: Okay. All right. Exciting. I wanna bring us to Recursive. So outside of our tangents, you have a pretty deep background in the NLP space. You worked on, like, early embeddings, GloVe with Chris Manning, who was a previous guest on the podcast, You.com. What's the history? How did you decide to start another company?From You.com to RecursiveRichard Socher [00:16:06]: Yeah. So I've been excited about AI for over 2 decades now. I sometimes feel like it's ancient history now. It's BC, the before ChatGPT era. No one cares about all the religions that happened, before, Jesus Christ, and no one cares about the models that happened before, transformers and ChatGPT and stuff. But, like, it's something that I've been deeply passionate about. I think AI is one of the most interesting things one could work on, period. I think language is the most interesting manifestation of human intelligence, too. And, at You.com, we eventually off-ramped from pushing, like the frontier of AI forward to mostly giving people, like, good search engines, search, APIs and answers over the web. I think that's an extremely important part of intelligence, just knowledge and access, especially even, we'll get there maybe later, if you wanna invent a eureka machine that invents everything for us, it needs to know how not to reinvent the wheel, proverbially speaking. And to know what has been invented, you gotta have internet access. So it's the number one used, most used tool, in LLMs, agents, chatbots, and so on is web search. So I'm really excited for You.com to own that and grow really well in that with really large customers and so on. But it's also not building frontier models anymore. And so I initially tried to do this within You.com and raise another round and so on, but you just can't. You have to do a certain thing, and until you print enough money that you're allowed to start a second thing within that company is really hard. At the same time, I had all these ideas. I put them into a book. I finished the book last year, and I was like, “It'd be really fun to work, on this myself.” I felt like with word vectors, and then prompt engineering and, ImageNet and larger language models for protein generation, not folding and so on, I, me and my teams have pushed the field truly forward. And I feel like we can do it again, here at Recursive. And in many ways, what I observed over the last, 20 years in AI is that whenever we replace some human part of the process of creating AI with a learned system, improvements follow. And so. We've done that taking out manual feature engineering, like in sentiment analysis. I don't know if you remember these old days where, like there are linguists, and they're like, “Here's how you negate, and there's a, like, regular expression.”Swyx [00:18:21]: I went to Penn where we — they had, like the WordNetRichard Socher [00:18:24]: That's right, WordNet, all of that stuff. YeahSwyx [00:18:26]: Original. They use, our grad students to label Wall Street Journal articles and, like, really construct a knowledge graph ofRichard Socher [00:18:32]: There you go.Richard Socher [00:18:33]: And WordNet started, was part of how we started ImageNet. But anyway, so, like, it was really, like, fun, to do. But when we replaced all of that manual feature engineering with vectors and neural nets and just backprop through everything, it started to work really well at scale. And so then everyone started to do architecture engineering, and I was like, “ that clearly can't be it.”Swyx [00:18:53]: You mean, neural architecture search?Richard Socher [00:18:55]: Like, manually, they would say like, “Oh, I'm, I'm doing sentiment analysis, so I have a special neural net that's really good at sentiment analysis.” And then the machine translation community had a special neural net for machine translation.Swyx [00:19:06]: I see.Richard Socher [00:19:07]: The summarization people had their own stuff. And I was like, “That clearly can't be it. We should unify all of that.” So I had 2 papers. One is called Ask Me Anything, and the other one was called DecaNLP. And DecaNLP eventually got cited, like, 5 times by the first GPT paper. And, to me, that was, like a really a big step forward. And then, of course, you had to combine this idea of prompt engineering with transformers and with language models, and you put it all together, you scale it up, which is also a huge amount of work. And then, the field progressed a lot. I feel like the next step and maybe the last step of that history and the arguably, success has a lot of parents, only failure is an orphan, like my version of that AI history, I do feel like in that history, you can think about, “Well, what's the next way to automate?” And that is the AI research itself, like the human, process of ideating, implementing, and validating ideas.Automating AI Research and Recursive Self-ImprovementRichard Socher [00:20:01]: And in our case, ideas for AI.Richard Socher [00:20:03]: And when you have AI then help you with that, it, by almost definition, becomes a self-improving AI ‘cause it now does research on itself. And there are lots of different misnomers. Some people think auto research is already recursive self-improvement. It'sSwyx [00:20:17]: Yeah, and you explained that in the talkRichard Socher [00:20:19]: Completely different.Richard Socher [00:20:19]: But, to me, it's the most interesting thing that I could be doing, and I'm really excited with the co-founding team. What's interesting is we have 8 co-founders in total, including myself. And soThe Recursive Founding Team and Darwin Gödel MachineSwyx [00:20:31]: They are gonna bring it up.Richard Socher [00:20:31]: Nice. Yeah. And they're all. I could talk about all of them if you want.Swyx [00:20:34]: Super stacked.Richard Socher [00:20:35]: Yeah. Just an incredibly talented group of people. And we all came to the same conclusion, but from very different directions. Like Josh Tobin, is our CTO. He ran, a bunch of different, projects at OpenAI, like, Codex and deep, research, agents and ChatGPT agents and so on. But before that, he also worked in robotics, and he saw the smaller simulations, and how it's gonna be really hard to scale that in full generality. And so that's, that was his angle coming to recursive self-improvement. We have Jeff Clune who's been working in, like, open-endedness for a long time, together with Tim Rocktäschel. Tim Rocktäschel also built Genie 1, 2, and 3, which is, like the most exciting and most sophisticated, I think, still world model, anywhere. And so they both came from this, open-endedness angle. Jeff also, I think, published one of the most exciting papers in recent years about recursive self-improvement called the Darwin Gödel Machine. Super interesting paper. If we could, maybe pull it up really quickRichard Socher [00:21:35]: It would be, like, super interesting to see ‘cause you seeSwyx [00:21:38]: By the way, I love how many paper citations.Swyx [00:21:40]: You're, you're giving people a lot of homework, which I like.Richard Socher [00:21:42]: Love it. Yeah. And so, like Caiming Xiong, a rockstar, we worked together at MetaMind and Salesforce Research together. Alexey Dosovitskiy invented the Vision Transformer, one of the most cited, papers in computer vision. Tim Shi is, like also a unicorn founder. Yuandong Tian led RL at Meta. So just like, yeah, really fun to work with them, and the next level of people are just incredibly strong, too. So it's been a really fun ride so far. So the first figure, you see exactly these kinds of ideas, that, I think, yeah, inspired a lot of us and now more and more people, where you have this archive of different coding agents. They learn how to self-modify, evaluate, and then create these phylogenetic trees, of, yeah, different ideas.Swyx [00:22:28]: That's one foundation. So that Darwin Gödel is an influence.Swyx [00:22:32]: Open-endedness is an influence. Any other trains of thought that feeds into Recursive that I'm missing?Influences: Open-Endedness and Learned SystemsRichard Socher [00:22:38]: Going to replace manual parts of the process of building AISwyx [00:22:42]: IRichard Socher [00:22:42]: More and moreRichard Socher [00:22:43]: With learned systems. Yeah.Swyx [00:22:45]: Which, and, like, merging different fields into one general, architecture.Richard Socher [00:22:51]: That's right.Swyx [00:22:51]: Okay. It seems like language models are already pretty generalist, right?Swyx [00:22:55]: Your next token predicting your reasoning. Was there a time that you thought, “Okay, these are good enough to have recursive self-improving machines”?Are Current LLMs Enough?Richard Socher [00:23:05]: It was clear to me that they will happen, within, like a year or two, and then it did exactly happen, like, earlier this year, right? Earlier this year, AI really went from not just being code, but being able to code. And that is a big unlock. It's definitely making everything a lot easier than it was, before the beginning of this year.Swyx [00:23:24]: One question that I think a lot of people have is the current LLM paradigm enough? Or, like, let's call it autoregressive transformer, with reasoning, whatever. Don't you need something else, some big unlock, whether it's world models, which Chris Manning is working on, or memory, continual learning, all that stuff? Or is it all of the kinds, and you think the current, let's call it transformer architecture, is here to stay and that's it?Richard Socher [00:23:48]: A lot of thoughts. So number one, I do think it would be great to have less of a monoculture in AI research.Richard Socher [00:23:55]: Like, if you look at, AI conferences now, I still remember the days in, like, 2010 when I tried to get my first neural net papers and NLP conferences accepted, and they just desk rejected them because, like, neural nets were something, quote, unquote, “We don't do in NLP conferences,” and just, like, desk rejected. And it was very brutal in the first years of my PhD. Now I feel like it's almost like the field switched to the other side. LikeRichard Socher [00:24:17]: Someone should try some other weird, crazy ideas now that aren't.Swyx [00:24:20]: There's also a few. I really respect, like, people still working on, like, GNNs and, like tabular stuff and.Richard Socher [00:24:25]: Yeah. Like, someone should still, like, do novel out there ideas. At the same time, I think whenever people say, “Oh, LLLMs are. Like, this is the end for LLLMs,” they just don't, like. LLLMs are also not the LLLMs of, like the past, right? Like, they are so much more sophisticated now. There's so many more clever things that people are doing. It — There's, like, different stages of training. You have the whole RL training, and you can take actions and, like all of these things where that can go really far. And then the folks that come from the neurosymbolic, direction say, “Oh, this will never work because they can't do neurosymbolic reasoning.” It's like, I think they're underestimating still the ability for these models to code, and code is neurosymbolic reasoning, and these models can code incredibly well. And so I do think there are, of course, more and more ideas that will be needed and we'll continue to have. We're seeing, like, more and more interesting high-level ideas coming out of the AI itself, too. And with really deeply integrating the fact that these models are code and can code, that line — I don't wanna give it all away, but, like, I think that line has a lot more to grow. But it's still an LLM, right? Even if that LLM codes for you and then runs that code in some integrated fashion. World models, I'm personally less bullish on. I think if you run a robotics company, you're gonna build your own world model. I think world models are super fun, and Tim Rocktäschel came to a similar conclusion after building the most interesting one with Genie 1, 2, and 3, which is gaming is a huge application for world models. Can see I sometimes got stuck in some games and, like, got a little overly competitive in the wrong direction. And so I understand games are fun, but personally, I'd rather work on science than gaming. And so, yeah, I think LLLMs, a lot more room to grow.Swyx [00:26:16]: Yeah. I think there's some interpretation of world models that some people have where it's like, well, it's okay, yes, there is that gaming element. There's this — there's the embodied robotics element. But the other part also is just, the more abstract sense of LLLMs are just modeling output, but they're not modeling the chain of thought, inside the human that has created the output. We can annotate it, of course, but, like, it's, it's always, like, this Plato's cave reflection of a thing rather than the thing, right?Richard Socher [00:26:43]: It's true.Richard Socher [00:26:44]: But I would argue that, and maybe we'll get there in the 10, spaces of intelligence, but I would argue that even our projection, our eyes is a projection of the real world. And, like, we have only a very narrow, band of the electromagnetic frequency spectrum that we can observe with our puny little 2 eyes and so on.Swyx [00:27:01]: It's good enough.Richard Socher [00:27:02]: It's, it's good enough for now, but, like the upper bounds of where it could be are so much higher. And, like, to map, the visual world the way humans see it is also not necessarily, like the end-all be-all for visual intelligence. And I would argue that language is still the most interesting manifestation of human intelligence. And while our visual cortex is certainly less sophisticated, than that of, certain animals all the way down to the mantis shrimp who can, have, like, 2 independent eyes, 3 bands, trinocular vision and each eye can see all the way to, like, floating temperatures in 4D and stuff.Richard Socher [00:27:36]: Like, mantis shrimp, you should look it up. It's likeSwyx [00:27:37]: Way OP.Richard Socher [00:27:38]: Super crazy.Swyx [00:27:39]: Yeah. ZeFrank, mantis shrimp.Swyx [00:27:41]: It's the best video in the world onRichard Socher [00:27:42]: I love ZeFrank, yeah.Richard Socher [00:27:44]: Big shout-out to him. But, like, I think there's a lot more room to grow, but none of these, other animals have language that's as sophisticated as ours, certainly not in writing. And once you can write, you can, start thinking about longer term civilizations. All of that is language. Programming is much closer to language. And I would argue, and this is, like an important thing in the spaces definition of intelligence also, is that all of these spaces are highly correlated, but visual intelligence is neither necessary nor sufficient for overall intelligence. You can be blind and still be an intelligent human being. And an AI can be blind and still be quite intelligent too.Swyx [00:28:25]: We were gonna bring thisRichard Socher [00:28:25]: Which doesn't mean that you're not more intelligent when you have it. Yeah.Swyx [00:28:28]: We're gonna bring this up. I might as well — Like, we have a classification of 10 types of intelligence that you had at the end of your talk. So I'm just gonna flash this up now for people to cover this. I don't know if, maybe we'll put this towards the end. We'll come back to this. I just wanna mention that, you do have a philosophy that I like when people do lists because then I can just go through this and then it gets — it's educational for people. But let's go back. I don't wanna get distracted. But, so effectively, I'll, I'll, reinterpret what you said as Yann LeCun is wrong. And then we'll justRichard Socher [00:28:56]: Don't quote me as that. I'm, I'm good friends with Yann. I think very highly of him in many directions.Swyx [00:29:01]: But he's wrong.Swyx [00:29:03]: You mentioned GPT-1, and I cannot let any, Alec Radford, mention escape. Did you talk with him when he was training GPT-1? Like, any historical, fun stories there that you might come up?DecaNLP, GPT History, and Scientific GatekeepingRichard Socher [00:29:18]: I did not, like, meet him a bunch of times. I think we met maybe once or twice at some conferences. But, like, he has told, I think Brian, the first author of the DecaNLP paper, that it did inspire him, and he cited it five times in the GPT-2 paper. So, and that's, likeSwyx [00:29:36]: Yeah, good enough.Richard Socher [00:29:36]: Very clearly said, like, this was the first instantiation where they showed in the DecaNLP paper, McCann et al, that you can just phrase every single NLP problem as here's some prompt, text context, here's a question and task description and here is some output. If you just do that enough, you can have one unified neural network model, which, by the way, also had all kinds of interesting attention mechanisms. There are slightly different formulations to the transformer. I think came out the same year, plus/minus a few months. And then you can unify all of natural language processing into one neural net. That is the core idea.Swyx [00:30:14]: And this was as opposed to at the time, LSTMs and what have you.Richard Socher [00:30:17]: LSTMs, but also, like, people being very stuck in thinking about one model per task. In factRichard Socher [00:30:25]: It's, it's kinda crazy, but the DecaNLP paper was publicly reviewed as, like, open, OpenReview. It was an ICLR submission. And, in it, you will see, how the whole community at the time thought about this. So, likeSwyx [00:30:43]: Some great contributions, but more work needed.Richard Socher [00:30:46]: So look at, like, search for not even for humans. Just scroll it up here. Like, question answering is not a unified phenomenon. There is no such thing as general question answering, not even for humans. And this is like, really, you replace your brain with a different brain a different neural net when you answer, like, different kinds of questions. It was unfathomable to the experts at the time that you can have one unified neural network that would answer all of these different questions. They are saying, “No, all of these questions require very different systems to answer, and trying to pretend they are the same doesn't help anyone solve any problems.” That's what it says right there, right? That's how hard it was to fathom. And now, of course, people, when I say, “Oh, we're gonna invent prompts,” people are like, “You can't even invent prompts.” It's such an obvious idea to have one neural network that, of course, does everything in NLP.Richard Socher [00:31:37]: But at the time, it was, like, extremely controversial, and the paper got rejected. And the sad thing is that it got rejected so hard and they were so certain that we stopped going on our list of things to try. And the number 2 or 3 on the list of extensions for this paper was add language modeling as another task. And then we could have, and that would have accelerated the timelines, in 2018, like, even further for humanity. But we got so crushed, and we were like, “Okay, maybe we'll just work on some of our other ideas for now and, like, come back to this later.” Yeah.Swyx [00:32:09]: How can we design a review system that rewards non-consensus?Richard Socher [00:32:14]: Honestly, I started to feel like arXiv is such a gift to humanity. With arXiv, you should just put your paper out there.Swyx [00:32:24]: Is it pre-preprints?Richard Socher [00:32:25]: Let — And honestly, I think Twitter X, people like you who pick up interesting papers, that is a better filter than the experts. Let everyone, like, have access. Now, of course, there are some downsides, which is, like, if you're super unfamous, you have no Twitter followingRichard Socher [00:32:41]: You don't wanna be on social media or whatever, you write a good paper, maybe someone, somehow no one notices it. But I would argue that if you just tell, like, 10 of your friends in your community about a paper and it is a really significant breakthrough, someone is bound to talk about it again. And, so I think science needs less gatekeeping. And, even though ICLR, with Yann LeCun, who started it, as one of the co-founders of ICLR back in the day, he also wanted less gatekeeping ‘cause he too was rejected for many years together with Yoshua Bengio and Geoff Hinton with all their early deep learning and neural net papers ‘cause it was just not the hot thing. And so ICLR started with that, but then it also started gatekeeping a little bit themselves on various ideas. So I think less gatekeeping, more open, and then allowing people to say, “Look, even if this is just on, or, quote, unquote, ‘just an archive,' if it has like 1000 citations, it's a legitimate paper. Doesn't really matter where you published it.”Swyx [00:33:34]: And I agree with that. I do think it's sad that I've heard that grad students have to do, like, how to Twitter, seminars to each otherSwyx [00:33:43]: Just because it's so important for publishing these days. This person is just reflecting the sentiment at the time.Richard Socher [00:33:49]: That's right.Swyx [00:33:49]: But it'sRichard Socher [00:33:50]: I think it'sSwyx [00:33:50]: It affected you so muchSwyx [00:33:52]: That you stopped work on it.Vibhu [00:33:53]: The sentiment also came out of some of the research, right? Like, the original BERT paper was trained, and towards the end of the paper, they're like, “Okay, throw off the last head, train specific iterations forVibhu [00:34:05]: Extractive summarization add a head for this.” Like, you should do task-specific stuff. These are, like the authors that wrote Attention, wrote BERT, telling you this is what you're meant to do. And, like the training tasks were also very odd. They're likeVibhu [00:34:16]: The — “We know that the model overfits to this weird mass language modeling. Throw away this part and just do specific models,”?Richard Socher [00:34:23]: Exactly. And, like, we had to try — come up with all clever ways of, like attention and pointers and so on to get the neural network to be able to do all of these tasks. And then some of them were better than state-of-the-art, some weren't, but we were like, “But it's still in one model.” I thought it was really cool. Really interesting.Swyx [00:34:38]: I was gonna move on next to Tim and open-endedness. He was head of open-endedness at Google.Open-Endedness, Rainbow Teaming, and Self-Set GoalsRichard Socher [00:34:42]: That's right.Swyx [00:34:43]: I don't know what that means.Swyx [00:34:44]: But he did a lot of talks.Richard Socher [00:34:45]: Genie 3 is one of the ways thatRichard Socher [00:34:47]: Rainbow teaming, yeah.Swyx [00:34:49]: So I first saw him at — speaking of ICLR, I first saw him at ICLR when he talked about open-endedness. He's he's done a few talks. Can we define what is open-endedness for people who have never been exposed to the problem? They are like, “What do you mean? I thought the only goal of AI is to optimize against a benchmark or.”Richard Socher [00:35:04]: That's right, yeah. It's a, it's a fuzzy term because there's so many different instantiations of open-ended, thinking. But, one way I often describe it, and certainly, Tim and Geoff Hinton would be even better at describing this, but it's a suite of methods that is more inspired by evolution than, very specific rewards. So in that sense, it thinks more about environments, about co-adaptation. And so a concrete example is in the cybersecurity and LM safety space where you have one LM that tries to attack another LM to say something unsafe.Swyx [00:35:40]: Yeah, the rainbow, yeah.Richard Socher [00:35:40]: And now the environment is the 2 having a conversation and now they co-adapting, right? They're like one makes a better attack than the first one inoculates itself somehow, like uses that as training data, makes it so it's harder to say something unsafe based on that. And then as the attack stops working, the attacker now tries a different angle, right?Richard Socher [00:36:00]: And that's why it's not just red teaming, but they're called rainbow teaming.Swyx [00:36:02]: So, like, don't tell me how to do things. Let me just figure it out myself.Richard Socher [00:36:05]: That's right. Think about the environments that you wanna use. Think about the rewards at a high level that you wanna, inspire towards, and then let the AI try out many more ideas in this interplay between sometimes humans, but also sometimes other AI agents.Swyx [00:36:22]: Yeah. I worked open-endedness into a model that I have been working on. It was the keynote for AI Engineer where you start. You, we have the token loop, we have the agent turns, and then we have goal. And I feel like the way that you're describing open-endedness is still somewhat of a goal. Like, please attack this,Swyx [00:36:41]: Other agent. But, to meRichard Socher [00:36:42]: Yeah, you set the rewards. You set the environments.Swyx [00:36:44]: The loop that makes the other loops is. What if the agent can set its own goals?Swyx [00:36:49]: And is it, is that open-endedness? Like, you don't give it a goal. Just, like, be a sentient being. And maybe sentient is a very loaded wordSwyx [00:36:57]: But just set your own directions. What do you think you should do?Metacognition, Subjective Goals, and Measuring IntelligenceRichard Socher [00:37:01]: I love this direction. I think this is one of the 10 spaces of intelligence, that I clump under metacognition and thinking about thought.Richard Socher [00:37:08]: And it's an interesting one. Whenever people say, “Oh, AI is like, this is, it's gonna stop from here. It's not gonna get that much better,” and blah, I'm like there's so many different spaces of intelligence that we haven't even started exploring yet and hence have made very little progress on. And there is an interesting, connection to economics and, capitalism. Like, it doesn't make sense for a company to build and spend billions of dollars building a model that instead of following the rewards and objective functions you gave it, may come up with its own objective functions and its own goals.Richard Socher [00:37:46]: Right? And then imagine you're like, “Okay, I spent billions of dollars. Now go develop this new battery, material for me and answer all my emails.” And it's like, “Nah, I think it'd be more interesting to evaluate the molecular composition of the atmosphere, on Jupiter.”Richard Socher [00:37:59]: And you're like, “That's not what I paid you billions of dollars for.” And so no one's working on that for good reasons. And then also, understandablySwyx [00:38:07]: It's not useful.Richard Socher [00:38:07]: It's not, it's not useful, and it could get a little bit weird, right? What if the AI does start to really have thoughts on its own, and what if we don't like those thoughts, right? And so it requires a whole different way of thinking about it. I had a great conversation with a good friend of mine, Sam Gershman, who's a neuroscience professor at Harvard, and, like, we just jammed on this a little bit on, like, what are the best meta goals. And, I do think, like, knowledge-seeking is a really good one. I'm currently thinking also about, like the ultimate measure and unit of intelligence broadly construed, and I finally have some. It's still too early to share it. It's not. I haven't fully baked the thoughts yet.Swyx [00:38:44]: Like some replacement for IQ.Richard Socher [00:38:46]: IQ is such a terrible definition, right?Swyx [00:38:48]: Elo.Richard Socher [00:38:48]: It makes no sense. Yeah, Elos are terrible, too, because it's always just like me versus others.Richard Socher [00:38:53]: But, like, you can be intelligent and not constantly compare yourself to others? And so, yeah, there's no, like. In fact, a lot of these definitions we have, which I briefly mention in my book, too, these definitions create sometimes explicit and sometimes a more implicit anthropic bounds. No dis to the company Anthropic, but just, like, this idea that your intelligence is like getting 100 out of 100 questions right on this IQ test. Well, if that's your definition then you can only be at 100 out of 100. Where do you go from there, right? So you see a lot of these, benchmarks that people are working on they, increase, they get close to human, maybe sometimesSwyx [00:39:30]: It's like an S-curveRichard Socher [00:39:30]: Slightly above human, and then it's flat.Richard Socher [00:39:32]: It's like, ‘cause that's your. If your definition is only that so tied to humans, you're only gonna get to just slightly better than that. So I think metacognition is a great example of that, where we're not even yet allowing the AI to think. We're not working on it very much, and hence there's very little progress in that.Profit Maximization, Real-World Environments, and Reward DesignSwyx [00:39:49]: Yeah. Well, we've interviewed Andon, which I think, has been working on the most open-ended, benchmarks, which is just real-world, money.Swyx [00:39:57]: Arguably, telling an AI to profit maximize is a bad idea.Swyx [00:40:03]: But they are doing it.Richard Socher [00:40:05]: I do think you don't want that super. Like, you don't want a superintelligence to have a ton of access to all kinds of tools and so on and then just give it that without some very careful reward engineering. ‘Cause it's like, I just buy a bunch of defense stocks and I start a war. I make money. Like, it's just like, it's a tricky situation, right? You just buy a bunch of stuff, short basic goods for people, and you create some weird famine, like, issues. Like, yeah, there's a lot of constraints you should put onto a trading system.Vibhu [00:40:35]: It's a fun measure, though, ‘cause, the bounds are very capped to where we're nowhere close to them. Like, in Andon Labs, the model's like, “Oh, it's Saturday, maybe I just close the store today.” “Someone's off. It's okay. We'll just close the store.”Swyx [00:40:51]: It's using Claude.Vibhu [00:40:52]: Yeah. ButRichard Socher [00:40:53]: Yeah, no. I'm not, I'm not arguing against it. Just, like as you get more and more intelligence, you wanna be more and more careful with that as, like an open environment, ‘cause the environment then is all of Earth.Applying RSI to Science and InventionSwyx [00:41:02]: Yeah. Okay. For recursive, not strictly necessary, right? Because, like, if your goal is you make a machine that, like, invents the other things, then, like, just solve, the science thingsRichard Socher [00:41:12]: Knowledge discovery, yeah.Swyx [00:41:13]: Solve machine learning research and discovery and all these things. Good enough.Richard Socher [00:41:16]: And eventually, so, our goal, I haven't really. I don't talk about it that often because it is a few years out, but our goal is once you have a recursive self-improving superintelligence, you then want to apply it to the most important problems. And I think a lot of those are in science and technology and broadly construed inventions, and those inventions in, physics to create better, cheaper energy with fission or fusion, in chemistry and to create better materials and better batteries and, better solar cells and so on. In biology, there's so much, like, I think soon to be low hang- lower and lower hanging fruit because of AI, because of protein and generation, not just folding, but generating new proteins like we did in ProGen many years ago. Like, so much positive impact we had if you take that superintelligence and you apply it to science.Swyx [00:42:04]: I do fundamentally believe that. There's a lot of approaches, though. You're not the only team trying and NeoLab trying.Swyx [00:42:09]: There's, like a lot of. Especially the physical sciences as well.Richard Socher [00:42:12]: And that's good. Yeah. I do think that physi- like the reason we are only doing it in a few years is that it's a little too early right now. Robotics is not quite there yet. The AI is not quite there yet. But I'm fairly confident in 3 to 5 years, all those constraints will be gone, and then applying to real physical robotics experiments and so on, like true robotic process automationRichard Socher [00:42:33]: Not the traditional RPA sense, but, like, having robots run experiments for you will be totally there. Yeah, it's gonna be great.Swyx [00:42:40]: Just to call back to something that you said early on about slow takeoff, you said that, like, while really the substrate that is limiting factor is, let's call this chips, and semiconductors and all these things, and you have race funding for that and, you are investing a lot on that. But have you done the math on, like, is it even- Achievable and, like, what is the, industry concentration needed in order to achieve, like, scale?Compute, Slow Takeoff, and Changing the Bitter Lesson SlopeRichard Socher [00:43:05]: Right now we know that, like, roughly, like a 1000 GPUs cost quite a lot of money.Richard Socher [00:43:11]: Right? If you wanted, like, 10s of thousands of GPUs, you're, you're talking billions and billions of dollars. If you say, like, one GB300 is, like, you could eventually create models that are, on that substrate, like are close and similar to human intelligence. And you want, like, thousands and thousands of, AIs to think about really hard problems, in a similar fashion to humanity. Like, yeah, that-that's, that's a lot of money. You do the math. It's like a lot. We don't have that amount of money right now anywhere to, like, build that. Now, things can get more efficient. You will have, I think, soon better algorithms that won't be, and better hardware that won't be as energy-hungry, and so on. Our human brain does quite a lot of flops with much less energy.Swyx [00:43:56]: 20 watts?Richard Socher [00:43:57]: That's exactly right. Yeah, that's the number often that's quoted. And, like, I think more, inventions will happen there, that then will accelerate the takeoff even further.Swyx [00:44:08]: One thing I always try to reconcile when talking, like, with new lab founders is, like, you're fighting Bitter Lesson all the time. You have to show initial progress, then you unlock the next tier of funding, then the next tier, then the next tier.Richard Socher [00:44:20]: Which unlocks larger model categories.Swyx [00:44:22]: Like, fundamentally, is that true? Like, are you fighting Bitter Lesson? Are you — will we have a way in which, like, no, we're changing the slope in some fundamentally different way?Richard Socher [00:44:31]: I do think we are changing the slopes in fundamental ways by making AI much more efficient, both in terms of the training as well as the inference.Richard Socher [00:44:43]: Yeah. I think we will — When you allow AI to do the work that it takes other labs thousands of people and years to do, I think we'll be able to get it down to weeks, and that will be much cheaperRichard Socher [00:44:53]: And hence, more affordable, accessible to others and so on.Swyx [00:44:57]: Yeah. You've shared initial results on that,Swyx [00:44:59]: Which, like, conveniently OpenAI has also done to their GPT-5.6, so we can talk about it now.Richard Socher [00:45:04]: Yeah. Yeah, so these areSwyx [00:45:06]: Let's recap what you've done.Early Recursive Results: NanoChat, NanoGPT, and SOL-ExecBenchRichard Socher [00:45:07]: Maybe, just a quick recap here. We built, this, system that isn't the full, even the full RSI system in its glory, but it is a first baby version of this. And then, we don't wanna just have it internally and not show anything and, just show some people of what's possible. And so we applied this to these 3 different tasks. One is NanoChat, by my friend Andrej Karpathy, just, like, train a small language model to get, really low bits per byte. And, like, hundreds if not thousands of people, used both their agents and themselves to try, to get to that, and then they got to 0.937. We literally took our system and got to a much lower, bits per byte, much faster within, like, I think less than 2 days. So we took this thing, applied our system to it, and less than 2 days later, we have — we outperformed every human and their agents, in, have ever worked on this. Same with NanoGPT. And then we're like, well, let's, apply it to something that's even more relevant, to real people and to the Nvidia ecosystem and applied it, to, SOL-ExecBench. And maybe you can scroll down to some of the, images. They're, they're kinda fun to see. But yeah, like, one you see has made some real inventions that weren't just hyperparameter tuning. Like, inventing hash tables and so on is quite clever. We have even better results now.Swyx [00:46:34]: What do you mean inventing hash ta — You didn't invent hash tables.Richard Socher [00:46:36]: Of course we didn't invent, like, hash tables. In the grand scheme of, like a hash table, it's like a super basic primitive in computer science. But to use it, for language modeling in this scenario inside a transformer and so on and to combine these ideas and put them together, that has then eventually also been invented, but there was a knowledge cutoff, and we did check that it didn't have access to that externally. We talk about this a little bit. If you scroll to the next figures, this is also an interesting one in that when you start from a really basic, poor, like, vanilla transformer, then we still outperform all of the community together. But if you start from the human seed from an expert like Andrej, then you get even lower. So the human seeds from which you start do still matter. So that was an interesting insight, in my eyes, on this. And then as you go, like, how long does it take to get to these models, to get to similar performance? It's much faster. And then a similar thing happens with the speed runs here where, people have worked on this for quite some time, and the model still was able to train a model more quickly. Why do we care about it? Well, speed of training is part of the equation of the cost, and ultimately, you wanna have the most intelligence per dollar, right? And so speed and quality are big parts of that. And, the,Swyx [00:48:00]: Yeah, the way I put it is, for people who don't understand they look at the chart, they're like, “Cool. What does it mean?” if you have, like a billion-dollar cluster and you can shave off 10%, that's 100 million dollars.Richard Socher [00:48:12]: That's exactly right.Swyx [00:48:13]: How much is that worth?Richard Socher [00:48:14]: Exactly. So when you click, when you look at, like the kernels, these kernels, yeah, for the non-experts, like these kernels are like, used in all the models. Every time you use an Nvidia GPU, you interface with that GPU through these kernels. And so here you see, the leaderboard best, and when it's recursive, and it's there are only a handful of kernels, in this whole benchmark where we weren't the best. And so to me, this is, like, really exciting, ‘cause it makes. It just showcases what this can do. And again these weren't like. We didn't, like, spend months or years, like, developing. In fact, in particular for kernel, CUDA kernels, like, we don't even have really deep. CUDA kernel experts in the team. And our system, that's the beauty. The system just did all of these things. We didn't invent this. And when we open source and release, things in the future and models in the future, like, it won't. They won't be the best in their, category or class or whatever because we're so smart, but it's because, we built a smart AI that does it for us.Reward Engineering and Good Auto ResearchVibhu [00:49:14]: Do you have anything that you've learned from how to guide good auto research? A lot of it also builds on human background, right? It's not just as simple as just, “Hey, go optimize this.”Vibhu [00:49:23]: But we do see it again and again, right? Like some of the Erdos problems, frontier math is being solved by people. And when they do a write-up, they're like, “Oh, I'm not a mathematician. I have no background in this?” “I saw some tools and I made it work.”Swyx [00:49:35]: While you're watching the World Cup, you're likeSwyx [00:49:37]: “This proves some conjectures that's going on.”Vibhu [00:49:40]: Yep. Any learnings fromRichard Socher [00:49:41]: Yeah, there's a Korean conjecture was. Yeah, that's pretty cool.Swyx [00:49:44]: To summarize, tips for good auto researchSwyx [00:49:46]: Versus bad auto research.Vibhu [00:49:48]: How did you build the recursive?Richard Socher [00:49:49]: Yeah. So without giving away all the secret sauce, maybe some things that are probably obvious to the experts but might still be interesting to some, folks is, like, reward engineering is one of the most crucial bits, especially, in order to avoid reward hacking. So you have to be really clever about avoiding. ‘Cause as your AI gets better and better, it will get better and better, at finding weird like, special cases or counterexamples and things like that. And so I'll give you an example. Like, when you ask to, like, make these 100, lines of code faster, and, how do you define fast? Well, you have one line at the beginning that says, “Start your stopwatch,” and one line at the end, “End the stopwatch,” and then, tell us how much time, progressed. And so, well, the simplest way is you just put that line that ends the stopwatch, rightVibhu [00:50:39]: At the startRichard Socher [00:50:40]: At the start. And then boom, it's now faster, right? So this isn't like this, like, super evil AI. It's just, like a very simple, dumb reward hack. And so you have to just very carefully think about all the different angles there. And then I think the longer time horizon the tasks are the harder it gets and the more interesting and clever you have to be to still use these kinds of ideas for it. But yeah, I can't give away too much there.Vibhu [00:51:05]: It seems like rubrics are taking a good spot in that, where for unverifiable domains, you have rubrics, you have a model breakdown, judge's criteria along the way.Swyx [00:51:14]: Yeah, it's a form of verificationSwyx [00:51:16]: Once you got enough rubrics.Richard Socher [00:51:17]: Yeah, everything. I said this a long time ago. That's why I've never been that impressed that AI can play games, ‘cause I'm like anything you can simulate and/or verify, you can have infinite training data forRichard Socher [00:51:29]: And hence, like, AI will solve it eventually.Swyx [00:51:32]: Looking for games where you can do auto domain distribution. So this is a game that nobody's trained on ‘cause it's a new game.Swyx [00:51:38]: And you can start gaming, you can start to play. So I've been building this and cloned this in person and it's just been self-play. I've had about a billion positions evaluated.Games, Self-Play, and the AI EconomistSwyx [00:51:48]: And, I wanted to do the AlphaGo thing of self-play until you ge

An Australian-led team has rebuilt thirty years of supernova observations into one consistent catalogue — and it adds fresh weight to the idea that dark energy is not constant. We also ask whether the young Sun swallowed a super-Earth, find out why a quiet Sun is bad news at thirty-five thousand feet, and watch JWST run the most sensitive exomoon search ever attempted. Plus a Crew-13 update and a skywatch with a live aurora alert for both hemispheres. In this episode ·         Two thousand, eight hundred and eighty-four Type Ia supernovae, rebuilt from Pantheon+ and the Dark Energy Survey's full five-year sample into a single internally consistent catalogue by Ryan Camilleri and Professor Tamara Davis at the University of Queensland, with ANU, Swinburne and international colleagues. ·         On their own the supernovae give a matter density of 0.310 for flat ΛCDM. Combined with the CMB and baryon acoustic oscillations, a tension appears under constant dark energy — and eases when dark energy is allowed to evolve. Preference: 2.5–3.1 sigma. ·         Why that is interesting but not a discovery, and why a second independent line of evidence pointing the same way as DESI changes the shape of the argument. ·         Professor Mutlu Yildiz (Ege University) on whether the young Sun engulfed a planet of 5–10 Earth masses — and whether that single event explains both the solar sound-speed discrepancy and the Sun's missing lithium. ·         Yaniv, Yair and Price on six balloon flights to 35 km: cosmic radiation at cruising altitude runs 40–60% higher at solar minimum, with an anticorrelation of r = −0.71 against solar activity. ·         David Kipping stacks twelve JWST transits of LP 890-9c and excludes moons down to 0.1 Earth radii across the entire Hill region — the most sensitive exomoon search on record. ·         Crew-13 now targeting no earlier than late September after an oxidiser leak in Dragon's propulsion system. ·         Skywatch: a coronal-hole stream arriving today with aurora chances at both ends of the planet, Venus at greatest brilliancy on 18 September, Mars past Pollux, Saturn towards opposition, International Observe the Moon Night on the 19th, and the equinox on the 22nd. Sources ·         University of Queensland — 'Big supernova dataset challenges dark energy theory', 8 September 2026 ·         Camilleri, Lee, Davis, Rubin, Shah, Scolnic, Lidman et al., 'Supernovae Unite: Combining Pantheon+ and DES-SN5YR', Publications of the Astronomical Society of Australia — arXiv:2609.05053; companion host-mass paper arXiv:2609.05321 ·         Royal Astronomical Society — ''Fingerprints' inside the Sun could reveal if it once swallowed a planet', 10 September 2026. Yildiz, MNRAS, DOI 10.1093/mnras/stag1527 ·         Yaniv, Yair & Price, Journal of Geophysical Research: Atmospheres, September 2026 — cosmic radiation at aviation altitudes across the solar cycle ·         Kipping, 'JWST Excludes Exomoons Down to 0.1 Earth Radii Around a Rocky, Temperate Exoplanet', arXiv:2609.05301, 4 September 2026 ·         NASA Space Station blog — 'NASA, SpaceX Adjust Crew-13 Launch Date', 29 August 2026; Canadian Space Agency update, September 2026 ·         EarthSky sun news and NOAA Space Weather Prediction Center outlooks, 12–14 September 2026 ·         NASA Science — 'What's Up: September 2026 Skywatching Tips'Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Today's episode — S05E190, Thursday September 10, 2026: Main story: 109 fast radio bursts have been used to measure how much galactic feedback has smoothed out the matter in the universe. A fast radio burst lasts about a millisecond, and on its way to us it is dispersed — low radio frequencies slowed more than high ones by free electrons along the path, exactly the way a prism separates white light. The size of that delay counts the ordinary matter the burst passed through. Writing in Nature Astronomy on 8 September, a team led by Kritti Sharma at Caltech, with Vikram Ravi, Liam Connor and Elisabeth Krause (University of Arizona), used 109 bursts localised to host galaxies out to redshift ~0.3 — most of them from Caltech's DSA-110 — to constrain the suppression of matter clustering by feedback, and the gas fractions of groups and clusters between 10^13 and 10^15 solar masses. They find more gas retained than the X-ray picture implies: gas fractions about 1.9σ above stacked eROSITA measurements, and slightly above the Atacama Cosmology Telescope's kSZ constraints, consistent with the bursts counting cool gas that X-rays miss. The constraint is already competitive with both of those far larger experiments, cutting the uncertainty on clustering at these scales by roughly a factor of eight. That matters because feedback is the confound sitting under the S8 tension — it mimics the signature of massive neutrinos and of some dark energy and dark matter models. The technique's foundation is Australian: the Macquart relation, established with ASKAP in Western Australia in 2020, is what found the universe's missing ordinary matter in the first place. The rest of the news: ·         Hypersoft X-ray sources: NASA and the Chandra X-ray Center announced a new class of object on 9 September — 84 sources across six galaxies (M31, M101 and four ellipticals) that emit more than eight times as many photons below 0.3 keV as just above it, exceeding 10^38 erg/s in that narrow band and more again in the extreme ultraviolet. Led by Mustafa Muhibullah (University of Alabama) with Jimmy Irwin and Rosanne Di Stefano (CfA), published in Nature Astronomy. They are likely compact objects accreting from companions, and may be both the long-sought progenitors of Type Ia supernovae and a significant unaccounted source of ionising radiation. ·         A rendezvous with Halley's Comet: a trajectory study circulated on 6 September by Roberto Flores and Elena Fantino (Khalifa University), Mauro Pontani (Sapienza), and Ivano Bertini and Cesare Barbieri (Padua) sets out the first Halley rendezvous achievable with proven hardware — unpowered Jupiter and Saturn gravity assists plus deep-space low-thrust arcs, a Hall-effect thruster on a standard RTG, ~2,000 kg launch mass with ~750 kg of instruments, launching August 2036 or September 2037 and arriving in 2060, about a year before Halley's 2061 perihelion. Months alongside the comet instead of Giotto's minutes at 68 km/s. ·         Apollo Maneuvers 2026: US Space Command announced on 8 September that it had completed the first live-fly orbital manoeuvre exercise of its kind, moving real operational satellites across low, medium and geosynchronous orbits with Operation Olympic Defender allies (Australia among them) and commercial operators via its Commercial Integration Cell. Named for the US Army's 1941 Louisiana Maneuvers. The practical consequence is more unpredictable objects in an already crowded region, and in-orbit refuelling becoming load-bearing. ·         Skywatch: New Moon on 11 September at 14:27 AEST (04:27 UTC) makes tonight and tomorrow the darkest nights of the month. Southern Hemisphere — the galactic core still high after dark, and Venus well placed low in the west-south-west, 7.4° from Spica and building to greatest brilliancy on 18 September at magnitude −4.8. Saturn rises about an hour after sunset for everyone, heading for its 4 October opposition with the rings about 7° from edge-on. North America — the pre-dawn sky, with Mars in Gemini passing Castor on the 12th and Pollux on the 17th, and Jupiter below it in Cancer. Plus a convoy of CMEs from AR4524 arriving since 8 September, two G1 storms already recorded and G2 possible: aurora chances at both ends of the planet, including Tasmania, coastal southern Victoria and New Zealand's South Island. Links & sources ·         Caltech — Fast Radio Bursts Poised to Help with Biggest Cosmic Mysteries (8 Sept 2026) — https://www.caltech.edu/about/news/fast-radio-bursts-poised-to-help-with-biggest-cosmic-mysteries ·         Sharma, Krause, Ravi, Connor et al., 'Signatures of Suppressed Matter Clustering revealed by Fast Radio Bursts', Nature Astronomy (8 Sept 2026) — https://arxiv.org/abs/2604.17162 ·         NASA — NASA's Chandra Unveils Mysterious X-ray Objects (9 Sept 2026) — https://science.nasa.gov/missions/chandra/nasas-chandra-unveils-mysterious-x-ray-objects/ ·         Muhibullah, Irwin & Di Stefano, 'Hypersoft X-ray Sources: A New Class of Luminous Cosmic Emitters', Nature Astronomy (9 Sept 2026) — https://arxiv.org/abs/2602.06192 ·         Flores, Beolchi, Pozzi, Pontani, Bertini, Barbieri & Fantino, 'Double Gravity-Assist Rendezvous Trajectory to Halley's Comet Using Deep-Space Low Thrust' (arXiv, 6 Sept 2026) — https://arxiv.org/abs/2609.02189 ·         Air & Space Forces Magazine — SPACECOM Maneuvers Satellites Across Orbits in New Exercise (9 Sept 2026) — https://www.airandspaceforces.com/space-command-maneuvers-satellites-across-multiple-orbits-in-new-exercise/ ·         EarthSky — Sun news: series of sun blasts arriving, auroras to come — https://earthsky.org/sun/sun-news-activity-solar-flare-cme-aurora-updates/ ·         EarthSky — Visible planets and night sky guide for September — https://earthsky.org/astronomy-essentials/visible-planets-tonight-mars-jupiter-venus-saturn-mercury/ ·         TheSkyLive — Moon calendar, September 2026 — https://theskylive.com/moon-calendar?year=2026&month=09Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Román Orús is one of the rare physicists who built a foundational mathematical tool — tensor networks — and then watched it become the engine of a unicorn. His 2013 introduction to tensor networks has been cited over 2,000 times; his company, Multiverse Computing, just announced a $570 million Series C at a $1.7 billion pre-money valuation. That arc — from condensed matter theory to Europe's largest quantum software company — is worth understanding on its own terms. But what makes this conversation particularly timely is a May 2026 paper Orús co-authored demonstrating that individual layers of Meta's Llama 3.1 8B language model can be encoded as quantum circuits and executed on IBM's 156-qubit Quantum System Two while the model generates text. It's a proof of concept, not a product — but it's a real result, and Orús is honest about what it does and doesn't prove.This episode is for listeners who want a technically grounded, hype-free account of the quantum-AI intersection: what tensor networks actually are, why they keep getting rediscovered across different fields, where classical simulation of quantum systems genuinely competes with quantum hardware, and what it looks like to build a company at the boundary between those two worlds.Sponsor MessageThe Capital of Quantum is a people story. Built on a top-five quantum PhD program  and 35-plus years of quantum research leadership. It's the billion-dollar initiative behind Discovery Center, launching this month with Microsoft, IQM, and Quantum Motion inside. That's why IonQ was born and is headquartered here, and why global companies keep choosing a spot minutes from Washington, D.C. This is where quantum is transforming the world. Come see it at the Quantum World Congress, September 23rd through 25th, College Park, Maryland. CapitalOfQuantum.com.What We Get IntoWhat tensor networks actually are — Orús explains the core idea without equations: tensors as the "DNA" of a quantum state, and how a network of them lets you see and quantify the internal correlations (entanglement) that matter versus the ones you can safely ignore.Why the same math keeps appearing in different fields — from condensed matter simulation to quantum computing simulation to machine learning, and why Orús sees that recurrence as a sign of something deep rather than a coincidence.How ChatGPT changed Multiverse's trajectory — the company was already applying tensor networks to machine learning before 2022; the emergence of large language models gave them a problem where the fit was obvious and the market was enormous.What "90–95% compression with minimal accuracy loss" actually means — Orús explains the overparameterization problem in current AI models and why he believes tensor networks address a genuine structural inefficiency, not just a tuning opportunity.The IBM kicked Ising model episode — Orús describes how his team rapidly produced a classical tensor network simulation of an experiment IBM had presented as evidence of quantum utility, and what that kind of competition between classical and quantum methods actually does for the field.The Cayley Unitary Adapter experiment — how Multiverse sliced individual layers out of Llama 3.1 8B, encoded them as quantum circuits, ran them on a 156-qubit IBM processor, and achieved a 1.4% perplexity improvement — and why Orús argues the improvement-per-parameter ratio is the number that matters, not the headline percentage.Why edge deployment is the real commercial driver — drones, satellites, vehicles, and industrial devices that cannot rely on cloud connectivity are the market pulling Multiverse toward smaller, more efficient models, not just benchmark competition with frontier labs.How Orús thinks about Multiverse's identity — he calls it a "quantum AI company," not a quantum company or an AI company, and explains what that distinction means for how they allocate research effort and where they expect to be when fault-tolerant quantum hardware matures.What he'd tell a PhD student today — a genuinely honest answer about the trade-offs between academic research and deep-tech industry, from someone who has lived both simultaneously.Resources & LinksGuest & CompanyRomán Orús — Personal Site — Lists talks, reviews, affiliations, and awards including the 2024 Physics, Innovation and Technology Prize from the Royal Spanish Society of Physics.Multiverse Computing — Official Website — Home page for CompactifAI, Singularity, and Multiverse's full product portfolio.Papers & Articles Discussed in This EpisodearXiv 2605.05914 — "Quantum-enhanced Large Language Models on Quantum Hardware via Cayley Unitary Adapters" (May 2026) — The paper at the center of the episode: Llama 3.1 8B layers running on IBM's 156-qubit Quantum System Two. Start here if you want the technical details behind the quantum-in-an-LLM result.Multiverse Computing — "Talking to a Quantum Computer" (May 2026) — The accessible blog-post version of the Cayley Unitary Adapter experiment; a good entry point before tackling the arXiv paper.arXiv 2401.14109 — CompactifAI: Extreme Compression of Large Language Models Using Quantum-Inspired Tensor Networks — The foundational peer-reviewed paper introducing Multiverse's tensor network compression method.arXiv 2509.06653 — "Classical Neural Networks on Quantum Devices via Tensor Network Disentanglers" (Sep 2025, rev. Apr 2026) — Orús, Singh, and Aizpurua on hybrid classical-quantum execution for bottleneck neural network layers; the research underpinning the longer-term hybrid architecture vision.Models & ProductsHyperNova 60B on Hugging Face — Open-source 50%-compressed model derived from GPT-OSS-120B; context for Multiverse's open model strategy.Pulsar 16B Launch with NVIDIA (June 2026) — Announcement of Multiverse's open reasoning model, scoring 87.22 on AIME 2025 at 16B parameters.Funding & Company ContextSeries C Announcement — GlobeNewswire (July 2026) — $570M / €500M round at $1.7B pre-money valuation; covers investor lineup and deployment verticals.Key Quotes & Insights> "We are using atomic bombs to kill a mosquito." Orús on the overparameterization of current large language models — and why he believes the transformer-attention paradigm, however successful, ...

A German rocket lifted off from an Arctic island on Saturday and became the first vehicle ever to reach orbit from Western European soil. Plus: the gaps in stellar streams may not be dark matter after all, two white dwarfs circling every 374 seconds in the far southern sky, peptides that fold in concentrated sulfuric acid, and what Katalyst's LINK is actually doing next to Swift. Both-hemispheres skywatch, and a serious safety warning for tomorrow's daylight Jupiter occultation.Links & sources ·         Isar Aerospace — mission updates — https://isaraerospace.com/mission-updates-overview ·         ESA — Spectrum launches to orbit (imagery) — https://www.esa.int/ESA_Multimedia/Images/2026/09/Spectrum_launches_to_orbit ·         NASASpaceflight — Isar 'Onward and Upward' — https://www.nasaspaceflight.com/2026/09/isar-onward-and-upward/ ·         Space.com — Private German rocket makes history — https://www.space.com/space-exploration/launches-spacecraft/isar-aerospace-second-launch-norway-andoya-spaceport-spectrum-rocket ·         Space.com — Gilmour Space eyes a second orbital attempt — https://www.space.com/space-exploration/launches-spacecraft/australias-gilmour-space-not-going-to-give-up-as-it-eyes-2nd-orbital-launch-attempt-in-2026 ·         Sky & Telescope — Gaps in stellar streams may be common — https://skyandtelescope.org/astronomy-news/gaps-in-stellar-streams-common-not-sign-dark-matter/ ·         Arora et al. — Semianalytic modeling of subhalo encounters with thin stellar streams, ApJ — https://iopscience.iop.org/article/10.3847/1538-4357/adf740 ·         Sharma et al. — Rapid orbital decay in eRASSU J060839.5−704014 (arXiv) — https://arxiv.org/abs/2608.09341 ·         Maitra et al. — eRASSU J060839.5−704014 discovery paper, A&A (2024) — https://www.aanda.org/articles/aa/ref/2024/03/aa47811-23/aa47811-23.html ·         MIT News — Peptides can form well-defined structures in Venus-like conditions — https://news.mit.edu/2026/study-peptides-can-form-well-defined-structures-harsh-venus-conditions-0831 ·         NASA Swift blog — Commercial spacecraft for Swift boost continues tech demo — https://science.nasa.gov/blogs/swift/2026/09/04/commercial-spacecraft-for-nasas-swift-boost-continues-tech-demo/ ·         NASA Swift blog — LINK spacecraft recovery progressing — https://science.nasa.gov/blogs/swift/2026/08/06/link-spacecraft-recovery-progressing-for-nasas-swift-boost/ ·         EarthSky — Saturn at opposition, 4 October 2026 — https://earthsky.org/astronomy-essentials/saturn-at-opposition-closest-brightest-best/ ·         Star Walk — Astronomical events, September 2026 — https://starwalk.space/en/news/night-sky-tonight-septemberBecome a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Piotr Lewandowski is a software engineer based in Poland who, as a side project, has built something that no well-funded research institute has: a continuously updated, ontology-tagged intelligence platform that ingests the entire quant-ph archive, tracks thousands of open quantum roles across hundreds of companies, parses patents and grants and open-source repositories, and links all of it to individual researcher profiles. He is not a tenured academic or a hardware engineer. He is an independent data practitioner, and that outsider position gives him a vantage point on the quantum workforce that insiders rarely have — or rarely share.This conversation matters now because the quantum industry is simultaneously claiming a generational workforce opportunity and struggling to fill highly specialized roles. Lewandowski's data offers a rare ground-truth check on both claims. If you work in quantum hiring, research, policy, or investment — or if you're a student trying to understand what the field actually looks like from the outside — this episode will give you a more honest picture than almost anything else currently available.What We Get IntoHow qubitsok's 500-tag ontology was built from scratch — why Lewandowski chose a tree-structured, parent-child tag system rather than relying on existing academic metadata infrastructure, and how it steers AI toward the most specific and useful classification rather than broad category labels.What the full quant-ph corpus reveals about researcher mobility — which countries are gaining quantum talent (Germany and China are notable winners) and which are losing it (the US and Australia are among the top brain-drain sources), based on tracking affiliation changes over time in published papers.The rising share of industry authorship in quantum research — industry-affiliated authors have grown from roughly 3.4% of quant-ph papers in 2005 to nearly 14% in 2026, and what that structural shift might mean for what gets published — and what doesn't.Why the platform tracks open-source contributions alongside papers — when researchers join industry and their publication rate drops, their open-source activity becomes a meaningful proxy for continued technical engagement, and qubitsok indexes both.The "qubie" talent-matching tool Lewandowski is building — rather than keyword overlap, qubie dispatches sub-agents to extract specific, claim-level evidence from a researcher's papers, dissertations, and other public writing, then returns a structured profile and interview guide for each candidate.Why sourcing quantum talent is a fundamentally different problem than general tech recruiting — the evidence of what a quantum researcher can actually do is largely public and published, but no one has had the infrastructure to read it systematically at scale until now.The two product directions Lewandowski is weighing — analytics and competitive intelligence for investors and companies versus talent matching for quantum hiring — and why he's currently prioritizing the latter.What it means to build a field-level intelligence platform as an outsider — Lewandowski is neither employed by a quantum company nor affiliated with a university, and that independence shapes both what he can see and what he can say.Resources & LinksGuest Linksqubitsok.com — The platform itself: quantum job board, daily arXiv paper digest with semantic tagging, and researcher collaboration search. All free for researchers and job seekers.qubitsok.com/collaborate — Search for quantum researchers by expertise, ontology tag, and affiliation — useful for finding collaborators or co-authors.Piotr Lewandowski on LinkedIn — The best place to reach him directly, especially if you're a company interested in early access to the qubie talent-matching product.Piotr Lewandowski on YouTube — His channel covering quantum computing job market analysis and platform updates.Papers & ReportsEPJ Quantum Technology — "The quantum technology job market: data-driven analysis of 3,641 job posts" (March 2026) — Lewandowski's peer-reviewed paper on the quantum job market, the most rigorous published treatment of the data underlying qubitsok.2026 Quantum Computing Salary Report — Analysis of 194 salary data points drawn from real job postings; a useful benchmark for researchers and hiring managers alike.Fastest-Growing Quantum Skills 2026 — Lewandowski's analysis of 1,212 quantum jobs and 14,975 papers identifying which technical skills are accelerating and which are declining in the labor market.Tools & Platformsqubitsok.com/jobs — Live quantum job listings, filtered by the platform's 500-tag ontology.qubitsok.com/region/europe — Live European quantum jobs market data, useful context for the EU talent and salary discussion.qubitsok.com/hire — Information for companies looking to use qubitsok's candidate database and, soon, the qubie matching tool.RecognitionDr. Bob Sutor's "Quantum Follow Friday" (February 27, 2026) — Sutor Group newsletter highlighting Lewandowski as a key quantum community voice worth following; useful context for his growing recognition as an independent analyst.Key Quotes & InsightsOn why the published record is a better hiring signal than a LinkedIn profile: > "Research gives you this unique lens to see people's work before you talk to them. You can be really prepared, and this helps on two sides — you talk to people actually capable of filling the role, and you're not wasting their time asking questions they already answered via their published work."On what brain-drain data actually shows: > "The biggest country that gained quantum computing talent in the last twenty-four months is Germany — which is not something someone could expect. And the biggest countries getting brain-drained are, interestingly, the United States."Insight — on the soul-crushing reality of quantum sourcing: Lewandowski's first job in college was sourcing — going through profiles with pen and paper, making cold calls that nobody wanted to receive. His argument is that quantum hiring doesn't have to work that way, because the evidence of what a researcher can do is already public. The problem has never been a lack of signal; it's been a lack of infrastructure to read it.On the rising share of industry authorship: Industry-affiliated authors have grown from roughly 3.4% of quant-ph papers in 2005 to nearly 14% in 2026 — a structural shift in who is producing the science, with implications for what gets published and what gets quietly redirected into proprietary pipelines.Insight — on the limits of the data: Lewandowski is consistently careful about w...

When we first dicsussed the Summer of Simulative AI in 2024 we knew it would be a brief summer, but it has recently come back with a vengeance with SimGym in April and now Simile AI's $2B Series B, backed by GreenOaks and Index Ventures with prominent backers like Fei-Fei Li and Andrej Karpathy, running tens of millions of simulations for Fortune 100 clients like CVS and 85–99% accuracy vs human focus groups. Time to catch up on why this Second Summer of simulation is working!From creating Smallville, the landmark 2023 paper on Generative Agents that showed AI characters could remember, plan, socialize, and develop emergent behaviors, to now building foundation models of human behavior, Joon Sung Park is trying to answer a much bigger question: what if we could simulate the world before making decisions in it? In this episode, the Simile co-founder and CEO joins us to unpack the path from generative agents to digital twins, why today's frontier models still fail to capture how humans actually behave, and what it would take to eventually simulate all 8 billion people on Earth.We go deep on Simile's approach to modeling human behavior: long-form interviews, observational and transaction data, randomized controlled trials, population-level and individual-level models, and post-training on the causal mechanisms behind why people make decisions. Joon explains how his research created digital twins that reproduced human behavior and attitudes 85% as accurately as people reproduced their own responses, why models optimized to be rational can be bad simulations of irrational humans, and why understanding “social physics” may require changing model weights rather than simply prompting frontier LLMs.We also explore the much larger ambition behind simulation: testing products and policies before deploying them, finding counterintuitive paths toward desired outcomes, modeling emergent behavior across entire societies, and potentially tackling problems like climate change, democratic instability, and UBI. Joon reflects on scaling laws for simulation, the economics of data-center-scale simulated worlds, the connection to Thomas Schelling and psychohistory, why simulation is surprisingly similar to painting, and whether we might already be living in one.We discuss:* How Smallville and Generative Agents led to Simile* Why Joon's team asked: “What if we can just recreate the world that we live in?”* Why useful personal agents require deep models of their users* Memory architectures, Markdown files, and the limits of prompting* “Social physics” and behavioral foundation models* Why web data captures what people say more than what they actually do* Interviews, transactions, observational data, and randomized controlled trials* Why predicting the future matters less than understanding how to shape it* How Simile creates representative simulated populations* Simulation versus prediction and the connection to Foundation's psychohistory* How to evaluate simulations instead of simply stacking LLM hallucinations* Creating digital twins of 1,000 real people and reaching 85% behavioral accuracy* Why frontier models can struggle to reproduce real human behavior* Why good simulations need to reproduce human biases and mistakes* Post-training models on randomized controlled trials* Population-level versus individual-level simulation* Scaling laws for human simulation* The long-term ambition to simulate all 8 billion people on Earth* Whether simulations could help solve climate change or detect collapsing democracy* Thomas Schelling and the history of agent-based modeling* Why future simulations could require an entire data center* Multi-agent simulations and what happens when simulated people interact* Replacing expensive human panels with synthetic populations* Why market research is only the starting point for simulation* Why Joon sees simulation as surprisingly similar to painting* Using simulation to study questions like UBI* Whether we are already living in a simulation* Why AGI and simulation may be the twin technologies of advanced civilizationsJoon Sung Park* LinkedIn: https://www.linkedin.com/in/joonspark* X: https://x.com/joon_s_pk* Website: https://www.joonsungpark.com* Simile: https://www.simile.comTimestamps00:00:00 Introduction and Joon's Path from Art to AI00:01:46 Smallville, Generative Agents, and the Origins of Simulation00:05:03 “Let's Just Create a World” and the Future of Personal Agents00:09:53 Social Physics and Behavioral Foundation Models00:14:08 Prediction vs. Simulation: How Do You Shape the Future?00:16:59 How Simile Models Real People and Populations00:25:35 Evaluating Simulations, Digital Twins, and 85% Accuracy00:30:23 Post-Training Models to Reproduce Human Behavior00:40:04 Scaling Laws and Simulating 8 Billion People00:43:10 From Schelling to Society-Scale Agent Simulations00:46:13 The Cost and Economics of Simulating the World00:52:05 Real-World Use Cases, Synthetic Populations, and the Market00:57:27 The Future of Simulation, Painting, and UBI01:04:23 Are We Already Living in a Simulation?01:06:08 Building Simile and HiringTranscriptIntroduction: Joon Sung Park, Simile, and the Story So FarVibhu [00:00:00]: Today, we have Joon in the podcast. Excited to kick this one off. Very exciting company. I wanna kick off and ask you the question, talk us through the story of your life. How have you gotten here?Joon [00:00:13]: Yeah, for sure. I'm really excited to be here. A story of my life. So I was born in Korea, and I lived there for a good 11 years or so of my life, and then my family moved to Boston. So we moved when I was 11, and my parents were doctors, so they were going through their postdoctoral studies. My dad was a surgeon, so he was doing his sabbatical years at the Boston Children's Hospital. So I grew up there, not too close to tech. I was very much a music and artsy, painting kind of guy.Vibhu [00:00:49]: Painting.Joon [00:00:49]: Exactly. I got into painting a little bit later, in high school, but that's what I used to do. And then I grew up mostly in the East Coast after Korea. So I lived a good number of years in New Hampshire, and then I went to college in Pennsylvania. And I got into more of this tech scene, in college. So I was originally trained to be an artist. I thought that would be my professional career. So it wasn't a hobby. It was like, “Hey, let's make a living out of this.” And then gradually, I got really interested in this idea of, hey, the greatest artist often creates their own medium, and the best medium that we had available today was in computation. So I decided to go deeper into that, and one thing led to another, and we can go deeper into this, but I decided that research was something that I gradually got interested in, and here I am.Smallville, Generative Agents, and the 2023 Breakout PaperSwyx [00:01:46]: So there's a lot that you packed into the research components. You had one of the best papers of 2023, which was the generative agents paper, commonly known as the Smallville paper.Swyx [00:01:58]: Feel free to call back to anything else that you mentioned, but most people would have heard of you from this. Do you have any statistics on how many people have, like, read it? arXiv gives you something, right? Some stats.Joon [00:02:10]: Yeah, it's a good question. How many people have read it, I'm not sure.Joon [00:02:14]: I know we do keep track of citations, and they are going up quite fast.Swyx [00:02:23]: Yeah, Google Scholar has 7,200 citations.Vibhu [00:02:25]: I feel like it made a bigger hit than that, and it was a pretty instrumental paper. It got cited so many times.Swyx [00:02:34]: It is frequently the answer when people ask, “What is the best paper you've read recently?” It's this one.Vibhu [00:02:39]: I thought the memory component was pretty underrated. It was a very good early memory system, and one of the biggest papers.Foundation Models and the Search for Killer ApplicationsJoon [00:02:47]: Yeah, so maybe I can talk a little bit about how this particular paper came together. So when I got into research, it was back in 2020 when I started my PhD program at Stanford, and that was the year, when we were about to get GPT-3 to be available. So we already had GPT-2, and you could sense that there was this new class of models that was just becoming available in the market, and the team got very intrigued. And the general consensus was, “Well, is this model going to be useful for anything?” “It's really strange that these models are not trained to do any particular task.” But we decided to take a bet. So a large group of scholars at Stanford, and it was led by one of my co-founders, Percy Liang, and we came togetherSwyx [00:03:35]: Who coined foundation models.Joon [00:03:36]: Who coined the term foundation models. We wrote this paper, where that term came from called Opportunities and Risks of Foundation Models. And during that process, really the thing that I started to think deeply about was, here is a model that is fundamentally new in our ecosystem. The reason why this was new was it wasn't, again, trained to do anything in particular, but its premise was it could do anything and everything. It was like a stem cell, if you were to take a biology analogy. And I got really interested in this idea that, well, if we were to really think about what are the killer applications that this particular technology would enable, what would that be? Many of my colleagues were using this for simple classification, simple generations. Interesting that these models can do that, but from an interaction perspective, not that interesting. We've known how to do that for many decades. And what we came down to was these models are trained on this very broad data from the web, right? So these are human behavioral data. It's social media, Wikipedia, all these data. So if you poke at the right angle, then you could see human behavior that would just pop out that's quite realistic, and we've never seen that before.The Time Machine Game and Recreating the WorldJoon [00:04:45]: So that got us really interested. The exercise that we decided to do, with this particular group of colleagues, Michael Bernstein, Percy Liang, and myself, who ended up becoming my co-founder at Simile, we sat down and we played this game that we call the time machine game.Joon [00:05:03]: Imagine we were to get on a time machine and fast-forward 10 years and look back. What would have been the single application that will have mattered that would be the most interesting and inspiring? And when we thought, “Well, what if we can just recreate the world that we live in?” it's really hard to get more ambitious than that. Like, let's just create a world.Joon [00:05:24]: And that's where we started. And initially, we had this paper that was a precursor to the generative agents paper called Social Simulacra.Swyx [00:05:32]: Before you go further, were there other candidates for the most ambitious thing in the time machine exercise? What was number two or number three?Personal Agents, User Models, and Why Simulation Came FirstJoon [00:05:44]: There is a close second that we were considering, which ended up becoming more of these automation tools, especially the vision around really personalized agents that would do things for you.Swyx [00:05:59]: That's also happening.Joon [00:06:00]: It's also happening. But it was interesting for us, right, in that the reason why, we decided to go with the idea of simulation, one, I was a huge science fiction nerd, and this idea of creating simulation, I was personally really just fascinated. I loved the idea. It's really cool to see, like, a game town like this and just see these agents live in it. But at the same time, my bet was if you were to create a really amazing personal assistant out of this technology, what you need first is an amazing model of your users. So I told a model, “Hey, can you go buy late dinner for me?” And it orders Hawaiian pizza, and I do not like pineapples on my pizza. Then it totally failed. The way for it to not make that mistake is only by having a deep understanding of who I am. And I gave a very simple and dumb example here, but you can imagine how this core understanding of people is instrumental. This is how, if we have our family and closest friends, they have a good mental model of who we are. That's the basis of our social connection. So our bet also was this technology around simulation, creating accurate representation of people ought to precede the more complex agents that would automate the world that we live in. So that was the bet. But that was a very close second, and I'm still very much fascinated by it. I think there's a lot of interesting work that's going around. My hot take here, though, is I don't think we've seen a true personal assistant that's useful, in ways that meet the ambition of that particular line of work. I think there are early applications that are interesting, and if you talk to even ChatGPT nowadays or Claude, they know a lot about us. So a lot of the generation it's doing, I do think it's much more tailored, but I think the ambition is quite large in that field, and I don't think we quite have all the right ingredients just yet.Swyx [00:08:01]: So OpenClaw and these personal agents, what do you want to see from them that they don't currently have?Memory, Markdown, and the Limits of PromptingJoon [00:08:09]: I do think it's slowly getting there, but I do generally want them to have much deeper understanding of the person. Right now, you look at the models. OpenClaw, what it's leveraging is a Markdown file, and I think it's quite clever, right? So if you look at the generative agents paper, this was the same intuition that we had, where initially when we were creating the memory architecture for the generative agents, and, like, this is, like, back in 2022, so we didn't really quite have the idea of even agentive architecture or the term agent. But the intuition that we shared with some of the work that's coming out today was we initially thought, “Well, do we want to make the memory into, let's say, knowledge graph? Do we want to train a bespoke model?” All of these things. And what we decided to do was, “No. Just forget about all this.” These language models are quite good at modeling text and understanding and reasoning about text. So just put everything in a Markdown file or a text file. You're done. I thought that was quite interesting that we could do that, and there's a lot of strength in doing that. But also, there are limitations. It's the way you retrieve and make sense of data that's extremely large, it takes a lot of work. So I think that technology is getting better. I also do, however, think, there are certain things you just cannot shape just by prompting the model. So to some degree, you do need to touch the parameters of the model itself. So there is this work that I do think does need to happen, and it is happening. The question is, how far can we take it? How do we source data, and how do you also create an ecosystem where people are continuously feeding data to this model so it's learning about you?Vibhu [00:09:50]: What's the intuition between why you need to do it in the model?Social Physics and Behavior Foundation ModelsJoon [00:09:53]: My intuition behind the actual when do you train or even post-train a model versus just prompt a model is if the model has to learn the underlying physics of the world that it's operating in. So it has to learn new social physics. The places where it doesn't have to train are the places where it already has the physics. We trust the physics. It already has the base statistics, but it's just trying to react to an environment. Then I think you can just prompt your way into getting the actions out of it. I don't think the models that are out in the open have yet learned the complete mapping of social physics of humanity. This is one of the core theses of Simile, right? And one of the core reasons why that is the case is if you look at the data that the model was trained on, these models were trained on the web data, like, whatever was available on the web. And these are really interesting data sets, but they are fundamentally the self-exposed attitudinal data with some behavior data that's sprinkled around here and there. And it has yet to learn the really deep behavioral nature of people, not just what people say they do online, but what they do in real life. And this is one of what I would consider to be the dark knowledge of humanity that we haven't quite captured. And it's these data that would also need to get factored into the model creation.Vibhu [00:11:21]: You call it behavior foundation model.Vibhu [00:11:23]: There's a good one-liner here, but outside of that, what type of data do you need? What are you changing on the model level? How do you go about modeling, doing a behavior foundation model?The Three Data Buckets: Interviews, Behavior, and CausalityJoon [00:11:35]: We think about data in three buckets. So one bucket is interview data. It's quite interesting. Rich qualitative data is interesting. It's not behavioral, but we would literally ask people, “Hey, tell me the story of your life.”Vibhu [00:11:53]: It's just what we're doing here exactly.Joon [00:11:54]: The question that you all asked at the beginning of this interview literally is the question we also ask. And we ask our participants to go a little bit deeper, than how far I went. Maybe I can give more of my life story in lieu of this. But the reason why that data is interesting is by learning about this very long-tail information about people, you get a lot of texture around this model, like, this person as a model. So even understanding their childhood memory or even their trauma, their first love, these things, quite informative in ways that's really hard to predict. So that's one. Then there are two tranches of what I would consider to be the behavioral data. One kind of behavioral data is observational. So these might be like transaction data, or these might be data that you can get by scraping the web, right? So you can imagine why these data sets would be interesting, right, because they give you the base statistics of people's behavior.Joon [00:12:55]: But then there is the last category of data, that I personally think is perhaps the most important, which is the data that describes the causal mechanism, the whys of people. Some of this is covered by the interview data, the qualitative, because people talk about why they made certain decisions. But really, where you get to see the most behavioral aspect of this is in randomized controlled trials, like RCTs. Imagine you have the same setup, but you have a few different variables that you are trying to tweak. Can you get realistic human behavior out of it in ways where, imagine you had this particular option. Imagine you're even trying to choose whether you're going to drink coffee or not. The day you drink coffee versus the day you didn't drink coffee, does your behavior change? That's a data set that describes a causal mechanism. This is quite important in modeling people. The reason why this is important is oftentimes when people come to us, or not just to us, but the reason why people are interested in simulation isn't because they want to predict the future. If you're trying to win against the stock market, predicting the future is interesting.Prediction vs. Simulation: Shaping the FutureJoon [00:14:08]: But most people, most decision-makers, what they want to know is, how can we shape the future? It doesn't really help you to hear that your sales are going to tank in two quarters. They're just gonna say, “Wow, that sucks.” What they want to know is, well, what do we need to do now to avoid that future? That's the causal mechanism. And this is also very hard data to come by, right, because the world is our ground truth, but it happens once. So in a very controlled setup where everything is equal except for one variable, this kind of data set rarely happens. So this is a reason why this data set is both hard to come by and quite important if you're trying to model human behavior.Swyx [00:14:50]: So behavior, I think, is the hardest data set to acquire. What is out there? What is even possible? You're not going to know a lot of details about my life. I don't even have data for myself on my own health or habits, and I just don't log everything. So how can you have that data?Joon [00:15:14]: So we run a lot of randomized controlled trials.Swyx [00:15:17]: But you put people in the lab, they watch them sleep, or what?Joon [00:15:20]: We do care a lot about the consent process. People know that we invite them to be a member of this community to both share data and have themselves represented in different forms. But we bring a lot of people to the lab, or virtual lab, where we design experiments that would pose them real behavioral decisions. And often in these experimental setups, what makes the difference between what is attitudinal versus behavioral is whether the stake in your decision is real. That's ultimately what makes it behavioral. So in these setups, we are inspired by our colleagues in social sciences, psychology, and so forth. So when they run studies, the techniques they utilize is imagine there's an online store that you're inviting people to come by. Then whatever they purchase in this experiment, they actually get that item delivered. Like, these are the things that make the stakes real. So we run a lot of these experiments, and we also do partner with firms. Right now, we also have customers who are quite excited to at least give us a glimpse of the behaviors that their users exhibit so that we can get a little bit deeper understanding of how people behave in these different platforms.How Customers Use Simile: Populations, Queries, and ExperimentsVibhu [00:16:39]: I think on the customer side, they have a lot of data about their users, who has bought. They have the action data.Vibhu [00:16:47]: Can you walk us through an example of what someone comes to you for? What questions would they want solved? Do you customize a model for them? Do you have something off the shelf? What does that look like?Joon [00:16:59]: Today, when people leverage our models, it's often to better understand the population of their interest. So usually, the start of the relationship, we come together and hear about what population they want us to model, right? So it might be that if you're a CPG company that's selling to all of the US, then maybe it's fairly straightforward. You want to model the gen pop of the US. But at the same time, if there is a vertical or if there's a market that they're trying to go into, imagine, they want to better understand, let's say, people in their 20s and 30s living in California. That's a much more specific population. So we hear about this population, and we go recruit these people, with consent, and with incentives, and we collect some of their data and create a model of these people. Then what our product allows you to do is query them. So it can take as input a filter that is a description of the population that you want to talk to, just like the one I just mentioned, and an environment. The environment can literally be survey questions, behavioral experiments, It can be A/B testing. Oftentimes, the core use cases are things like concept testing, to start with. But also, people sometimes want to do focus groups or one of the fun use cases that we also serve is even modeling things like earnings calls for public companies.Joon [00:18:21]: So these are the use cases that we often start with.Swyx [00:18:23]: Concept testing, is that an established term? I've never heard of concept testing.Concept Testing, Gallup, and PoliticsJoon [00:18:27]: Yeah. So it has to do with they have, let's say, different messaging, different products, different ideas.Swyx [00:18:32]: It's like a marketing exercise.Swyx [00:18:33]: Okay, got it. Got it. Politics?Joon [00:18:36]: We do, have a strategic partnership with Gallup, and of course, Gallup is deep into policy space and so forth. Right now, we have not worked deeply with politics, like that area just yet, however.Swyx [00:18:49]: I'm curious if there is demand or if they really would have different needs that somehow fundamentally don't mix with your existing, users or people.Joon [00:19:00]: I think there's certainly demand.Joon [00:19:02]: But we are very much mindful of how this technology gets adopted and the societal impact that we'll end up having with this technology. And I do see politics as an area where a company has to be particularly thoughtful about the way they operate and make impact. So this is where we also want to make sure that we form enough of guardrail and perspective on how to leverage this technology before we go on to serve markets like the politics.Swyx [00:19:29]: I'll give people an example. one of my favorite shows is The West Wing. I don't know if people have watched.Swyx [00:19:34]: One of the key storylines is, like, the president has, multiple sclerosis, but they haven't. they need to figure out how to disclose it. So they run a poll with a fake governor and ask people to respond on the poll,Counterfactuals, Polling, and When Simulation Is UsefulSwyx [00:19:47]: They try to make decisions based on the results of that poll on, like, how well they'll be received, like where, how should we play this?Swyx [00:19:54]: And I'm like, well, I think those counterfactual things, I would use a simulation for this if I could trust it.Joon [00:20:01]: For sure.Joon [00:20:02]: In that show, how'd it go?Swyx [00:20:04]: In that show, it was, like a foregone conclusion. They were like, “We know it's bad. We just don't know how bad.” And then the poll came back. It was like, “It's really bad.” And then they just did it anyway.Joon [00:20:14]: Part of it is to show, right? So you're, you're looking at the ideaSwyx [00:20:17]: Maximizing drama.Joon [00:20:18]: How bad could it be? Oh, it's horrible.Swyx [00:20:20]: And to some extent, I think that is part of the trick of the, or the challenge or with being a customer of yours, which is that if I know it's. if I roughly know and can intuitSwyx [00:20:35]: What the effect is going to be, do I need you? What sensitivity of it, of effect do I need in order to make a decision, right? So for example, if I, my approval rating is 50%Swyx [00:20:48]: And I, they have this negative piece, news item comes out, and it drops to 30.Swyx [00:20:52]: If it drops to 20, if it drops to 40, do I care? No. It, I know it drops. It's negative. So when do I care about simulations?Joon [00:21:01]: You do something that's clearly bad, that's not popular, and people don't like you, like, yeah, it's likeSwyx [00:21:05]: You don't need a simulation.Joon [00:21:07]: Yeah. Well, so there are a couple of things. one is, there are use cases where, like every day, developers, designers, policymakers, marketers, every single day, they create assets. They create new products. And turns out, it's many of the decisions in hindsight is obvious. Yes, of course this is bad, but we still run those studies because understanding the magnitude and understanding how acute something is quite difficult, even if, we feel like, of course, like this makes sense. this is the reason why we make so many mistakes. Like, every time somebody goes online and say something that has huge backlash, you look at that and like, “What an idiot.” However, it's tough. That's one. There's also another aspect here, which is, again, this is the reason why simulation is different from prediction. In simulation, in the ideal case scenario. So what simulation is trying to show is it's trying to show each step of the way or each step that we need to take to get to a certain outcome, right? So in the most advanced simulations, sometimes the next step that we're suggesting might be quite counterintuitive. The analogy that I sometimes give, and I ground it in a more realistic example, but, I, as I mentioned, I'm a huge fan of science fiction, and I don't know how, many of the audience members have read, like, things like the Foundation series by Asimov.Simulation as a Path, Not Just a PredictionSwyx [00:22:37]: Oh, yeah. We've mentioned psychohistory a number of times.Joon [00:22:39]: Okay, fantastic. So I might be, talking to the right crew. If you read Foundation series, literally the first act is there's a group of scientists who have found out that, “Oh, our galactic empire is going to collapse, and we're going to have 30,000 years of unrest.” And they run psychohistory, the simulator that tries to teach them, “Okay, how can we keep this unrest to a 1,000 years?” And they plan this out, and the first step of that plan is to get the scientists who say, “Okay, this is coming,” exiled into this random place in this, galax- galaxy.Swyx [00:23:18]: Terminus.Joon [00:23:19]: Exactly. And that's so counterintuitive. Like, what a strange move that you literally sent the group of scientists who was raising voice around this potential collapse of galactic empire into nowhere. How is that the right first move? Well, it turns out in this particular simulation, that was the move.Joon [00:23:40]: It's these things, right? And the reason why these reasoning is possible is because you're showing the step function or each step that results in a particular outcome. So really what simulation allows you to do in its highest form is you give it not a problem or question, like what would people answer to the survey? That's not what we do. What we tell it is, “Here is a goal that we have. In the context of foundation, we want to keep the unrest to a 1,000 years. What is the path that we need to take now to get to that particular future?” And that's what simulation allows you to do. Now, translating that into real market, imagine you're a automobile company and you're about to release a, EV, and you're trying to understand, well, how do we market EV, to make sure that our stock price goes up? But what if the answer comes down that, well, you can market your EV in XYZ way, but that might change people's perception around the cars that's not EV and make your overall sales to go down. Not very intuitive, especially all you're trying to optimize is EV salesss, and that's the only thing that you're tracking, then that might result in a completely wrong solution, or at least different solution than what you would have expected, whether it's right or wrong.Joon [00:24:57]: That's the power of simulation.Swyx [00:24:58]: For listeners, we covered a similar topic with Mikhail Parakhin from Shopify, where they are working on SimGym. I don't know if he ever talked to you about it. it's very similar.Joon [00:25:07]: ISwyx [00:25:07]: The goal is increased conversion, but then the journey is very unusual.Joon [00:25:12]: Journey is unusual.Swyx [00:25:12]: Yeah. The-- He's trying to look for interventions on a shopping trajectory, which is similar to what you're saying. Like, it's not about the attitudinal, is your word for it.Swyx [00:25:24]: It's about behavior.Joon [00:25:25]: It's about behavior.Swyx [00:25:25]: And that's exactly the difference, right? It's, like, not about the near-term direction about-- but it's more about, like, how do you affect multiple turns of interactions.Vibhu [00:25:35]: You had a good quote at the start about this as well. It's not about people wanting to know the outcome. It's about how they can change it, change the way to get there, something like that. But I wanna take it back to how do we know this is grounded? LikeGrounding and Evaluating Digital TwinsVibhu [00:25:47]: How do you run evals? How do you test that simulations come through? if I was to do the same thing that you described with, say, your favorite LLM, Opus, GPT-5.6, have some agent to map out these thingsVibhu [00:26:02]: How different are the answers we would get if I give it the same goal, the same objective, make a decent system? You're saying that you need to change the model weight. You have your own solution to this. But how far off are we, and how do you check if it's grounded? you have some interesting stuff on your site that points to how you run real evals, but if you could take us through that side. I think that's one of the big concerns that people have. They're like, “LLMs hallucinate.”Vibhu [00:26:27]: “You're just hallucinating layer after layer,” right?Joon [00:26:30]: The way we do this, and this is the paper that we worked on after the generative agents paper that really became the, at least for Simile and also the field of simulation and synthetic panels, really became the foundation. Yeah, this is the paper. the paper is called Generative Agent Simulations of 1000 People. Here's what we've done. For this paper, we brought 1,000 people that's representatively sampled from the US to a virtual lab. And what we have done was we spent two hours collecting fairly wide-ranging data. In this particular study, we focused a lot on this interview data, that was, whose script was taken from this project called American Voices Project. And then we would also pair that with a lot of behavior data and so forth, whatever we can collect within two hours. And then we would send these people away for a couple of weeks. And during that time, I would use this data to create their digital twins. And I would bring the humans, participants back after 2 weeks and have them complete a battery of surveys, experiments, behavior studies. So we have the list here, which included things like behavioral economics games. We would run literally, like, Big Five personality test, General Social Survey. We would also go ahead and run the randomized controlled trials that were published on PNAS. And we would have their digital twins predict how the source individuals would have acted in these studies and surveys. And this is where we could replicate people's behaviors and attitudes 85 percent as accurately as people would replicate their own. So that was the first really paper that gave this validated results that we can model individuals in an accurate way. And what we ended up finding now, of course, in AI space, so this paper came out at the end of 2024. AI space, a year and a half, 2 years, that's a lifetime.85% Accuracy and Why Frontier Models Miss Human BehaviorSwyx [00:28:24]: Yeah. Just, for listeners who are not seeing the YouTube, I just wanna say, like, the headline figure is 85 percent accuracy, like, which is a big improvement over all the otherSwyx [00:28:34]: Methods that you showed.Joon [00:28:36]: But the part that was particularly striking to us, especially as we improved this technology even further, was the generative AI models like ChatGPT, Claude that's coming out, it does give you the right foundation. However, what they do not consider is the true attitudinal and behavioral aspect of people, especially in the population that you care about. So what these models are really good at today is they're trying to become the super rational, objective machines, right? So you go get their data from places like Mercor, Scale. You talk to professional programmers, scientists to create model that's amazing at reasoning. That's what they do. Simile doesn't care about any of this. The models that we're talking about here, what we're trying to create are models that are as dumb as I am, right? So if I make some mistakes, the model has to make the same mistake.Swyx [00:29:34]: Oh, that's very hard.Joon [00:29:35]: That's very hard.Swyx [00:29:36]: You're solving Murphy's paradox.Joon [00:29:37]: That's exactly. And this is a completely different data and training objective. This is also where we see quite a bit of discrepancy in the performance in human behavior prediction between the frontier models, Simile's model, and the models being created in this space, where in some cases, the model performance of frontier models go all the way down to 20, 30 percent, especially if you go into that more niche population on topics that our customers would care about. On more gen pop, it might be around 50 to 60 percent. So it's not very robust. Like, you wouldn't want to make your decision off of these and these findings. If you can bring that up to 85 percent, that is ultimately what people end up getting very excited about.Swyx [00:30:20]: Yeah. Do we wanna keep going on the paper, routes?Joon [00:30:23]: Yeah, for sure. So the last one, was an interesting one. So this, paper was the follow-up paper that we had, to the 1000 agents paper, where the idea was now can we augment the models even further and post-train a model based on a lot of randomized controlled trials? So this was an interesting one. The data is always the most interesting part of modeling in many ways. The data that we got here was there's this, there's this platform called Open Science Framework. So some, the audience might be familiar with this. And there has been, especially in the social sciences over the past 5 years or so, there has been this concern around replicability of studies. And so it was a bit of a crisis, the scientists acknowledged, where we rerun the study and we don't see the same finding.Post-Training on RCTs and Replication StudiesVibhu [00:31:12]: Oof.Joon [00:31:12]: It's tough. And the reason why it's there-- that was often the case was there's this survival bias where the papers that get published often need to maintain what we call the value of less than 0.05 in the experiments that we ran. That suggests that only-- there's only 5% chance that the results that we saw is false positive. But the tricky part was all the papers that were not published, and there's still a 5% chance that whatever we publish is totally just randomly generated. Like, there's a 5% chance that, hey, this effect is not real, but it just happened to be real because of the sampling bias. So because of that, what scientists started to do was they started to register their studies. So before running an experiment, they would go to this platform and say, “Here is the data. Here is the population that we're collecting, and here's the hypotheses.” And they would just say, “Here is our hypothesis.” Like, “This is what we believe.” And you cannot retroactively change those hypotheses. This is what gives us more scientific statistical confidence that whatever effect that you ended up seeing is true. So that ended up creating this really interesting platform where there's one platform that has now contains tens of thousands of real-world experiments and hypotheses. And a lot of these are really high-quality, like, professionally designed behavior studies and random- randomized controlled trials. So we got the data and the studies from this platform and used that to make a point. And this particular, model is not, something that we're serving commercially because this was a part of the open science. But this particular data set, helped us make a point that by collecting a lot of these randomized controlled trials, that are really well-designed, we can make significant improvement in model's capability to predict human behaviors. So that's what this paper was about.Vibhu [00:33:10]: Is this stuff done on a individual level? Like, do I need to tune the model per individual, per company? Is there foundation model changes and then some slight post-training? Anything you can share there?Population-Level vs. Individual-Level ModelsJoon [00:33:21]: So this particular model was trained. the data we had at the level of individuals, but this particular model was trained. We experimented with both. And this is what we end up doing at Simile too. We always train 2, distinct model. One is what we call the population-level model. The other is what we call the individual-level model. And both take very similar input, which is the description of a subpopulation or individual and a stimuli. In this particular work, we've done the same. Here, the results that we are reporting are much more geared towards individuals because we do think that is a harder task in many ways, but that's what we have done.Vibhu [00:34:02]: You seen anything on the questions that humans can solve that models can't solve? So likeHuman Biases, Mundane Choices, and What Models MissVibhu [00:34:09]: Currently, it's, I live 5 minutes walk away from a car wash. It's a 10-minute drive. Should I walk or drive?Joon [00:34:16]: Huh.Vibhu [00:34:16]: The model will say, “Oh, walk to the car wash.” And, you don't have your car.Vibhu [00:34:20]: Is anything like this a problem in simulation? You would assume, like, very simple for human to think about, but if the model is saying you should walk to the car wash, anything here?Joon [00:34:32]: It's less, what can we solve, but I think it's more about what biases or mistakes do people make that models miss. Like, imagine that you are, like the. When I was still at Stanford, I lived in Palo Alto. So it's about, I would say, 40-minute walk from the campus. You ask the model, “Okay, let's go home. What can I, what can I do?” It would likely call an Uber or, give me, the bus time. But for the longest time, I really liked walking back. And the reason why I wanted to do that was not for efficiency. It really helped me think. And I like to walk for, half an hour or 40 minutes or so a day, where I just get to, just think about ideas, research, just get lost in my thoughts. That's very human activity. Unless the model has seen that and understands the importance of that activity, it would miss these kinds of features. So that I think, is fundamentally what we're trying to model. Like, what is fundamentally human might not be the most efficient thing to do, might not be the right thing to do, but things that make us who we are.Swyx [00:35:43]: I'm curious if, there are some data sets that you really want that would materially help you. One version of this may be interesting, which is more valuable to you to acquire as a data set, all of LinkedIn, all of Twitter, all of Facebook?What Data Matters: Social Media, Transactions, and FacebookJoon [00:35:57]: It's a little bit hard to rank, in part because, there's, there's this product saying where no feedback is wrong because it teaches you something about your users. Doesn't matter what feedback.Joon [00:36:11]: I think it's a little bit like that.Swyx [00:36:12]: So just whatever is bigger.Vibhu [00:36:13]: What about a different domain? Say it was. What about all of Amazon data?Joon [00:36:17]: Oh, yeah.Vibhu [00:36:18]: Shopping data, right?Joon [00:36:18]: Shopping data. So Amazon data is interesting in that it's very much behavioral, although, like, what people do on social media, you could squint and say that is also behavioral. But the transaction data is always interesting. It is also most commonly available, however.Joon [00:36:33]: If we were to look at purely social media, like if you really, if I were, if I had to really pick, Facebook likely is interesting because I do think it is most a default version of people. Because you go to LinkedIn, it's very much professional environment. So people put up their, they have their guards up, right? And that still is interesting because that is true human attitude and behavior, but it is not your base state. you go to Twitter- Twitter, people have their own crazy personas, or depending on who you are. Like, my Twitter profile and, persona is very much, initially was I was very much an academic. “Hey, I'm here to share my studies.” Now, I share, things that's related to Simile. But Facebook is one of those more private space where people just connect with their friends. In that way, I do think it shows you a little bit more about who that person is. So if I had to pick, I'd likely pick, Facebook.Swyx [00:37:30]: Yeah. And you're interested in, like, the whole person and their background and philosophy. I, is it too clinical or too machine learning-oriented to just say this is just ways to inject variance and biases? The broad question, is, like, is this any better than a randomized, like, combinatorial explosion version? So we have a link to the TencentBillion Personas, Synthetic Demographics, and Bespoke DataSwyx [00:37:54]: Billion persona paper, where they did not do any of the groundwork that you are doing.Swyx [00:37:59]: They just did like a cross matrix of here's all the professions in the world, here's all the people, possible backgrounds in the world, do a dot product across all of them, and that's it. That's your prompt for a billion people.Swyx [00:38:12]: This will do something. I don't know if it'll do what you do, but it gets you some way, some percent of the way there.Joon [00:38:18]: So this was an interesting paper. Like, what I admired about this paper when it came out was the scale. And you do gradually want to be able to simulate really large societies and interactions. So the scale is definitely admirable. it is relying heavily on the known statistics that went into training the model. So to the extent that you believe that statistics is correct, this is not a bad way to go about this. But the thesis here, and this is something that we also have seen in the market, like if this works, then we have solved simulation.Joon [00:38:54]: It,Swyx [00:38:55]: Because I survey, like, okay, 5% of the US population is in construction.Swyx [00:39:01]: The other 5% is in medicine, whatever, right? And then you just keep going down the list, and then you do the other side. 5% has, like, the big 5 personalitySwyx [00:39:08]: Of, like, neurotic or whatever. That's it.Joon [00:39:11]: That's it. So if you believe that the underlying data set and the platform that we're leveraging has all the right statistics, then this will have solved it. you're at that point merely retrieving the knowledge that is already embedded in the model, in the model parameters. That's not, unfortunately, what we see, where there is such detailed and also niche knowledge about people that if you just take one example, it might feel very mundane, but it's quite rich when you put together, that you do need to do a lot of bespoke data collection to better understand people. And this is also, I think what makes this particular, job fun, which you want to deeply understand people, and the process of deeply understanding them requires a lot of attention to the details. And you do need to pay attention to and pay respect to the daily lives that people lead.Scaling Simulation: From Thousands to SocietiesVibhu [00:40:04]: I wanna talk about scaling simulation.Vibhu [00:40:07]: So what can't we simulate, what can we simulate, and how does scaling affect this? So how big are the models? What if we go from, 8B, like, couple 100 billionVibhu [00:40:18]: Like billion000 parameters, billion000? Do we get scaling? Any interesting emergence? Like, at a certain scale, at a certain amount of training, you uncover anything unusual and any learnings from that?Joon [00:40:31]: What we are seeing is at Simile, so we do post-train our own model. The thing that we're seeing is the early glimpse of scaling law in simulations. The more data about humans and more compute you ingest, you start to get predictive and predictable gains of the model performance in simulating it, simulating people.Vibhu [00:40:51]: Ooh. We need a scaling law curve.Joon [00:40:52]: It's scaling law. Whenever you find it's a beautiful thing. And we're starting to see the glimpse of it, which is quite exciting. But if you talk about the ambition of simulation as a whole, it's not merely about building a model. It's about building a model, then creating the agents that become the individuals in a much larger ecosystem. So they're creating this multi-agent simulation. Down the line, you want these multi-agent simulation to also live in a very rich environment, right? What we are really trying to get to at that point is, hey, can we create. All right, let's do a time machine game again, and 5 years, 10 years into the future, can we create a simulation of 8 billion people living on Earth? I think that's quite interesting. And that really is the vision. And once you get to that state, the questions that you can help answer for the society also start to change from my perspective. The answers are fundamentally about emergence of the emergent behavior of society and large groups of people.Joon [00:41:53]: So the questions that I get excited by, and maybe this is a stodgy- a bit. I have my, academic side of me.Joon [00:42:01]: And for me, it's questions like, can we help solve climate change? If you look at climate change as a problem space, this is what we, like social scientists would often call it the wicked problems, problem where you have many actors with competing incentives for trying to make a very complex decision and coordinating that coordination decision. Very difficult to really solve in real life, which is also the reason why we couldn't solve it. Can simulation help us solve that? Another one is, can we understand the signals for collapsing democracy, or can we understand or can we uncover the origin story of the monetary system? These are societal questions that we never really had a good way of answering. If we can create simulations of our society, you have to believe that these are the problems that we can solve. So that's really the ambition of this field. And, I also think, yes, I think there's a Nobel Prize to be won there, which wouldn't be surprising. And I think there's some amazing societal impact that we can have to help people make better decisions.Climate Change, Democracy, and Societal SimulationSwyx [00:43:04]: Nobel Prize in economics?Joon [00:43:06]: In economics.Swyx [00:43:06]: Oh, I see. I see. Rooting for you to write that paper.Joon [00:43:10]: One of these days. But, one of the scholars that I was deeply inspired by, When I was coming into the space of simulation, is this scholar, named Thomas Schelling.Schelling, Agent-Based Models, and the Nobel PrizeSwyx [00:43:23]: Schelling point?Joon [00:43:24]: So the canonical example of the work that he's done was he was one of the creators of agent-based modeling. So this was, like, in the 1970s and 80s. It's very early days, but this was truly one of the first exemplars of simulations. And one of the canonical model from that time, and of course many of these simulations are trying to tackle the societal problems that's most relevant for their era, it was called the model of segregation. So racial segregation was a big topic, that, we cared about. And what they've done was they created this grid world where they had red dots and blue dots. And these dots were, back in the day, like, they were the agents, and they had a simple rule that governed their behavior. If certain percentage of your neighbors are of different color and if that goes above certain threshold, then you move to a new location at random.Joon [00:44:21]: One of the striking finding of this paper or this agent-based model was for the longest time, people thought the segregation within society was caused by explicit and overt racism.Joon [00:44:34]: But if you look at this model, people's preference towards living with people of the same color, that preference can be very minute.Joon [00:44:42]: But the very small difference causes the society to segregate completely over time. This was very counterintuitive for a lot of people. And this particular work ended up informing housing policies. Mixed income housing, got really inspired by this work. And Thomas Schelling ends up winning the Nobel Prize for having laid the groundwork for very early versions of simulations. The opportunity that I do see here in the more scientific terms, is agent-based models for the longest, had impact in the 1980s, 90s, to some extent, early 2000s, but it has now gotten forgotten by the community a little bit. Because as you can imagine, red dots and blue dots is not really a rich description of people.Joon [00:45:31]: But with the emergence of things like generative AI and, in particular, generative agents, we do have an opportunity to create these agent-based models that are high fidelity enough to help us make really complex decisions. And that's the opportunity that I see. If that truly works, then yes, that is the work that will result in a Nobel Prize.Swyx [00:45:53]: Yeah. For what it's worth, and I grew up in Singapore. 80% of Singapore is in public housing, and public housing has, enforced racial quotas for exactly that reason, which is very interesting. okay, so we talk about scaling, we talk about all these, the agent possible applications.Cost, Reuse, and the Economics of SimulationSwyx [00:46:13]: I'm scared about the cost. if you even-- let's just keep it to the US, about 8 billion people.Swyx [00:46:21]: But, how much does it cost to model so many hundreds of millions of people?Joon [00:46:26]: Oftentimes today, we don't start at that scale, this stage of the, of industry and simulation as technology. But we can get our users extremely rich and meaningful insights even by modeling thousands, tens of thousands of people. And today what we do is every week we are collecting data on the scale of tens of thousands people's data, and we have panel partnerships that gets us to tens of millions of people globally. So that's what we do today.Swyx [00:46:55]: And just as a side note once you've collected one person for one studySwyx [00:46:59]: Can you reuse that same person for all the subsequent studies?Joon [00:47:03]: That's exactly right.Swyx [00:47:03]: Okay.Joon [00:47:04]: The beauty of this model and these agents is the fact that they are domain-agnostic.Joon [00:47:08]: That what you're really trying to understand is what is the fundamental nature of these people? What's their social physics? And there are a lot of, a lot of, people that does change over time. Like, even, like, even things like, how many times have you gone have you been to, like, CVS the past week? that will change. But there's so many traits about people that are also known to never change. Like, your risk tolerance doesn't really change over time. It's very consistent. So it's these things that we're trying to learn. But the scale we are operating is right now hundreds or, tens of thousands to hundreds of thousands. And in many of the core use cases that we are deployed in, and this is more than enough population, to cover those. Really, at that point, what you care about is less the number of people, but more do you have the right subpopulation of interest covered? And this is also the reason why people want a larger sample. It's not because they want, stronger statistical guarantees. It's more that can they filter down to any population of their interest. However, you can also imagine in 10 years, if we truly believe that the compute is going to scale, that we'll have much more availability for compute, and our ambition for simulation is also going to scale accordingly, there's definitely a reason for us to create an entire data center worth of simulations.Joon [00:48:35]: Or in my hunch here is I do think in the next some number of years, we will start creating simulations that will cost as much as training a foundation model. But perhaps it's going to be so valuable to the society that it would be a no-brainer. Right now, even today, like, we are training bunch of new foundation model just so we can say we trained one and we spent tens of millions. But if we can create a simulation at the level of society that would solve climate change, I would run that today. I would raise the money right now just to run that.Multi-Agent Simulation and Social InfluenceSwyx [00:49:10]: Amazing. the follow-up question is, does it also compound if you let the simulations talk to each other?Swyx [00:49:18]: Or do they already do that today? They don't, right, as far as I understand?Joon [00:49:22]: It depends on what simulation you're trying to run.Joon [00:49:24]: In the multi-agent simulation setup, the agents do talk to each other.Swyx [00:49:28]: Right, which is exactly Smallville, right?Joon [00:49:29]: That's right.Swyx [00:49:30]: But a lot of times, for example, in commerce, you're just by yourself, so there's no point talking. which is way cheaper.Vibhu [00:49:37]: But they use all these levels, right? Like, you decide what you will buy based on what other people around you buy and talk about, right?Swyx [00:49:43]: It depends.Vibhu [00:49:44]: It depends.Swyx [00:49:45]: Again, I'm, I'm coming at this from a cost point of view. I'm like, “Oh my God.” LikeVibhu [00:49:48]: I thinkSwyx [00:49:49]: If there is, like, some combinatorial thing of, like, thousands of people talking to thousands of people, then that one million X's might cost.Vibhu [00:49:56]: I have a very different view as the cost point aside. Like, running these studies in reality is a lot more expensive, right? Running any study like this is you gotta have people do it, you gotta sign people up. It's very expensive and sometimes, like, not feasible to run the study.Vibhu [00:50:14]: But the outcome or the decisions you make are very expensive on them, right? So spend X million on something that, the overall process costs 100 million might as well, right? There's, there's a lot of value to be had there. It's a small cost, but I'm excited on the cost side.Joon [00:50:33]: To some extent, and when you deploy technology, you often want to deploy in a way where you can replace existing budget or you can make things more efficient, and that is the best way to deploy. However, the way you capture the long-term value of the technology is making the argument that, no, it's the upside, that by making this better decision using simulation, you have saved yourself or made yourself hundreds of millions or even billions of dollars, and that's a case to be made.Vibhu [00:51:06]: Random tangent question. So if you're doing a lot of inference, a lot of model multi-agent stuff, are you at the point where it makes sense to, train a model that' very sparse? You're expecting to do multi-million dollar runs. Are you thinking about this in model architecture standpoint or inference efficiency, or, you're still at the research phase of it works, we're not super there yet?Joon [00:51:34]: Efficiency, we do think quite a bit about. this is technology that is deployed now in some of the largest enterprise companies in the world, and we do process significant number of queries, that are trying to, simulate the populations in the world. So efficiency is a consistent thing. we don't want to over-optimize too early, so I wouldn't say, like, this is the higher bid Right now, but this is definitely something that we think pretty carefully about.Swyx [00:52:05]: Yeah. Are there other case studies? So we, you talked about CVS, talked about Gallup, Deloitte, Wealthfront.Efficiency, Enterprise Use, and Real-World Case StudiesJoon [00:52:12]: Wealthfront is an interesting one, because one of the things they were trying to do, they were one of the first customers that wanted to do product testing that goes beyond just asking people what they think about, let's say, behavior experiments and so forth. So there, really what we had to do was reason about multimodal input, so images, but also you can also imagine, like, these agents traversing through Figma mockups or websites. So some of the things that our agents can also do is it can be given a domain, like, or, like, a website URL and go use it for a while. It's these things. And Wealthfront was one of the first, customers, that was very excited about this possibility.Vibhu [00:52:53]: What have people been asking? Like, is there any demand that we have not covered? Like, UI testing, right?Vibhu [00:52:59]: I wanna try a new. I wanna ship a new feature, test the UI, simulate how people will do it. Any interesting things that you're seeing demand for?Product Testing, Websites, and Synthetic PanelsJoon [00:53:08]: Today, a lot of the demand does come from like, the places where people have historically used human panels, we can now replace with agents, and these synthetic populations. And this is not replacing human panel. in many ways, the simulation that Simile is building is grounded. So the way that I think about this is we are trying to represent humanity at scale. And in that way, the use cases are what we would expect, but it's the scale of deployment that surprises me.Joon [00:53:44]: Turns out there are so many decisions that people make every day in these organizations, groups, and we want to be able to say, “We listen to people. We have consulted our users.” But in reality, that is rarely the case because getting to people and asking them many questions, it's difficult. It's both costly, time-consuming, but most importantly, people are just not available. If I had to answer 1000 survey questions for this one particular, vendor, even if I wanted to do that, like, I would never do it. And that's very much the case. What simulation can do is ensure that the voices of people are always represented in rooms where the decisions for them is made, right? So all the stakeholders of this particular product launch, ideally they're consulted. That's what this technology really is trying to enable.Market Size, TAM, and Human Decision-MakingSwyx [00:54:39]: In my mind, that means it skews towards more consumer focus, right? Like, anything with a wide enough customer base where you do benefit from the diversity that you represent. What are some rough statistics, just for people who are not familiar with this market in general, what's the market size that. I'm sure you have some, like, rough numbers. market size is, like, a vague questionSwyx [00:55:01]: But, like, how much do people spend?Joon [00:55:03]: So market research is a $100 billion industry.Joon [00:55:06]: But the thing about simulation is not a tool for market research. Simulation is a tool for human decision-making. So the question around what is a TAM here is quite tricky, right? Because it's easy to say, “Well, market research TAM is roughly 100 million or 100 billion.” so is it a TAM? And not really, right? Because in many ways, you're trying to inform all human decision-making. You're trying to inform every decision that are made about humans for humans. What is a TAM for that? It's really unclear. And I'll be honest. Like, I have a scientific background, I have a research background, so I didn't come into the field calculating, oh, what is the TAM for human decision-making? But I just had to assume, well, if we can inform every decision that is made about human for human, that has to be big.Swyx [00:55:58]: Some- something valuable.Joon [00:55:59]: Exactly.Swyx [00:55:59]: To some extent, you are a unicorn founder now, and you have to care as a CEO. But, like, I do think, like, yeah, when you go into these boardrooms with people that you're quoting millions of dollars of contracts for, like, you have to say, “Well, here's what you spend on humans-”Swyx [00:56:15]: “. And here's what we save you, and it's 85% similar.”Joon [00:56:19]: And certainly, the value case, is something that we care deeply about. Like, what is the value that we provide to the users and the decision-makers? But this is also where, like, as a founder, I think valuation only tells one very superficial aspect of the story, and I try not to think too much about valuation, in general, because that's not what also motivates a team or certainly doesn't. I'm, I-- Again, the interesting thing about researchers is we are happy living in academia, getting paid next to. we get paid okay. we don't get paid that much, as a researcher here in academia, but it's the impact and it's the, it's the value that we can provide to the individuals and the society that really drives us. And in that way, ultimately what drives us is the impact. Does the simulation we provide have a real impact in people's decision-making in ways that progresses our society forward? If the answer is yes, then yes. that has to be great business, and we see that in numbers, and we do care deeply about that upside story, but that's the heart of it.Where Simulation Goes NextVibhu [00:57:27]: Do you have any timeline predictions? So we talked about scaling laws of simulations.Vibhu [00:57:33]: You brought up, okay, maybe one day we can simulate how to solve climate change.Vibhu [00:57:38]: Where are we now?Vibhu [00:57:40]: If that's not the end state, what is an end state, and what does progress look like?Joon [00:57:45]: So what I sometimes tell people is simulation as industry, it feels a lot like where GPT-3.5, GPT-4 was, for the AGI saga, which is we have now technology that is powerful enough to do real damage on the verticals that we are tackling. At the same time, there's a lot of progress that is yet to come. And that's, I think, where this is. So the way I see it, I do think there will continue to be breakthroughs both in data, in algorithms, and there will be much more aggressive scaling that will also happen over the next few years. But I think that's roughly where we are.Swyx [00:58:27]: I think that was about the ro

In this episode, we meet the CEO of one of the most important institutions in mathematics - arXiV. Around since 1991, the website is a publicly accessible home for mathematical research before it is officially published in journals. Around 1,000 articles are published on arXiv every day — it has hosted everything from the proof of the Poincaré Conjecture to some of the newest research from AI companies like DeepSeek and Gemini.  We spoke to Ramin Zabih, a Professor at Cornell Tech and Interim Chief Executive Officer at ArXiv, about the organisation and how it supports the very best of mathematical research. Ramin explains why early access is so important, the site's role in making maths more robust, and why arXiv can publish things that journals can't. Particularly importantly, he talks about the impact of AI on arXiv's work.  To learn more: Watch a recording of Ramin speaking at the INI What is the Poincaré conjecture?   Another approach to the current state of artificial intelligence in mathematics  All of our content on artificial intelligence  Maths on the Move, is the podcast from plus.maths.org. We speak to researchers from the frontiers of mathematical science so you can connect with the maths that shapes and explains our world. Hosted by Plus editors Rachel Thomas and Marianne Freiberger.

Modelli di linguaggio di grandi dimensioni e programmazione agentica nel mondo dei beni culturali. AI generativa, ontologie e archeologia aumentata: come si insegna a una macchina a capire la cultura.Dallo studio distribuito di digitalia:Franco Solerio, Francesco Facconi, John DoeProduttori esecutivi:Sponsor:Squarespace.com - utilizzate il codice coupon "DIGITALIA" per avere il 10% di sconto sul costo del primo acquisto.Links:On the Dangers of Stochastic Parrots — Bender, Gebru, McMillan-Major, Mitchell (2021)Stochastic Parrots — voce WikipediaRetrieval-Augmented Generation (RAG) — voce WikipediaCos'è il RAG — spiegazione AWSSPARQL Query Language for RDF — specifica W3CSPARQL — voce WikipediaKnowledge Graphs Generation from Cultural Heritage Texts — paper arXiv (metodologia ATR4CH)Knowledge Graphs vs. Large Language Models: Competitors or Partners in Supporting Virtual Museums — ACM JOCCHKnowledge Graphs for Cultural Heritage and Digital Humanities — Proceedings ACMDocStoLab — Università degli Studi di SalernoLangGraph — pagina ufficiale LangChainLangGraph — repository GitHubGoogle Agent Development Kit (ADK) — documentazione ufficialeADK — annuncio Google Developers Blogn8n — piattaforma di workflow automation con AIIntroducing the Model Context Protocol — annuncio AnthropicMCP — specifica ufficialeModel Context Protocol — voce WikipediaIntroducing gpt-oss — annuncio OpenAIgpt-oss-120b — model card su Hugging Facegpt-oss — repository GitHubClaude Code — pagina ufficiale AnthropicClaude Code — repository GitHubVibe Coding — voce WikipediaSpec-Driven Development with AI — GitHub BlogHow to write a good spec for AI agents — Addy OsmaniAI Agent Orchestration Patterns — Microsoft Azure Architecture CenterLangGraph Multi-Agent Workflows — blog LangChainGingilli del giorno:Supporta Digitalia, diventa produttore esecutivo.

Photo by Zac Taheriyan on Unsplash Published 10 August 2026 e564 with Michael, Andy and Michael – a spider filled episode with AdTech and lifesaving robotic implications, VR Mario Kart, quantum lawn mowing, identity management and a whole lot more! Michael, Andy and Michael begin the episode with Spider-Man showing up in an unexpected place – a BMW infotainment system!  The negative reactions from the articles and cohosts were not surprising.  The team have discussed ads you don't want showing up where you don't want them – like on automotive windshields, in AR, and more – see e481 from 2024 in the links below for more on this. The team shifts to a more enjoyable driving experience, where Andy describes what he did to enable a VR version of Mario Kart.  Next up – a reflection on 30 years of Quake with an anniversary release.  Then, appealing to the team's sensibilities on mowing lawns and penchant for NP quantum problems, a lawn mowing optimization game from Pudding.   Switching from software to hardware, the cohosts take a look at a surprising smaller than expected device called the Octopus 16 which measures EEG brain signals to allow the user to control a computer.  See the show notes below for the original box for the Emotiv headset from the July 2015 kickstarter. Then, the titular post for this show, a new robot based on nature for rescuing people who are drowning.  The inspiration for this robot was a fishing spider who walks on water to capture its prey.  Your show notes thankfully do not include pictures of such spiders, though if you click through to the Arxiv paper, there is a picture, so consider yourself warned.  While not discussed in this episode, Dr. Kate Darling's book The New Breed: What Our History with Animals Reveals about Our Future with Robots explores social and ethical themes on robotics that fit nicely into this theme.  Michael M had a chance to hear her speak at Duke earlier this year and an article about that talk is included in the notes below. Rounding out the show, the team wraps up with an identity service from Meta called Facebook Verified.  This free service is intended to help identify when a person is a human and distinguish from AI agents.  The cohosts shared thoughts about this topic and the service before the recording time was exhausted. Aside from spiders, what other robotic form factors based on the animal and insect kingdom would you want to see?  Have your bots

Kalorien, Makros und Aktivitätsdaten versprechen Kontrolle – doch wie verlässlich sind Tracking-Apps, Portionsangaben und die Kalorien der Smartwatch wirklich? Wir erklären, für wen Ernährungstracking sinnvoll sein kann, welche typischen Fehler die Ergebnisse verfälschen, wann der Zahlenfokus problematisch wird und wie man das Tracking später wieder ausschleicht. ------------------------------------------------------------------------ Dominiks Buch zur pflanzenbasierten Sporternährung im UTB-Verlag: https://www.utb.de/doi/book/10.36198/9783838560328 Dominiks Gesundheitscommunity: www.gsundes-hannover.de Dominiks Online-Knie-Kurs: https://gsundes-hannover.de/knieschmerzen/ Dominiks Online-Rücken-Kurs: https://copecart.com/products/34bd5abb/checkout Marcs veganes Online-Fitness-Coaching: https://vegainer-academy.com/ Marcs Online-Kurs: https://www.copecart.com/products/a50f88f2/checkout ------------------------------------------------------------------------ Werbung: Dieser Podcast wird unterstützt von der Firma Watson Nutrition. Die Firma bietet als einzige umfassend laborgeprüfte Nahrungsergänzungsmittel für eine optimierte Nährstoffversorgung. Zum Angebot zählen Multi-Supplemente, Mono-Supplemente, Sportsupplemente wie Kreatin oder auch Proteinriegel, Shakes und essenzielle Aminosäuren Mit dem Code veganperformance erhältst du 5 % Rabatt auf deine Bestellung.  Zur Firmenwebseite: Watson Nutrition ------------------------------------------------------------------------ Redaktionelle Anmerkungen 00:03:01–00:03:25 und 00:11:31–00:12:07 Einordnung: Die genannten FFMI-Werte sind nicht als harte genetische Grenzen einzelner Personen zu verstehen. Sie beschreiben statistisch modellierte Bereiche, die in den untersuchten Populationen nur von einem kleinen Anteil erreicht wurden. Aus den Daten lässt sich daher nicht ableiten, dass ein bestimmter Mensch einen höheren Wert unabhängig von Training, Ernährung und Zeit grundsätzlich niemals erreichen kann. 00:55:20–00:55:43 und 00:57:09–00:57:28 Einordnung: Ein tatsächlich bestehendes und über längere Zeit anhaltendes Energiedefizit führt zum Abbau gespeicherter Körperenergie. Aus den von einer Tracking-App angezeigten Zahlen lässt sich jedoch nicht sicher schließen, dass dieses Defizit auch wirklich besteht. Fehler bei Portionsgrößen, Lebensmitteldaten und der Schätzung des Energieverbrauchs sind möglich. Kurzfristig kann der Gewichtsverlauf außerdem durch Wasserhaushalt, Verdauungsinhalt, Zyklus, Medikamente oder Erkrankungen überlagert werden. 00:59:09–00:59:51 Einordnung: Dass eine niedrigere Energiezufuhr nach einiger Zeit leichter fällt, belegt nicht automatisch die Entwicklung eines neuen biologischen „Setpoints“. Ebenso denkbar sind Anpassungen der Gewohnheiten, der Mahlzeitenzusammensetzung, der Appetitwahrnehmung und des Umgangs mit Hunger. Die Regulation des Körpergewichts ist komplex und lässt sich nicht auf einen einzelnen festen Sollwert reduzieren. 01:11:55–01:12:30 Einordnung: Mikronährstoffwerte aus Tracking-Apps sind nicht grundsätzlich wertlos, häufig aber deutlich unvollständiger und unsicherer als Angaben zu Energie und Makronährstoffen. Fehlende Datenbankwerte bedeuten nicht, dass ein Lebensmittel den betreffenden Nährstoff nicht enthält. Die Angaben können Hinweise auf mögliche Ernährungsmuster geben, eignen sich aber weder zur sicheren Beurteilung des Versorgungsstatus noch zur Diagnose eines Mangels. 01:26:25–01:26:53 Einordnung: Das Beispiel einer möglichen Nutzung von Apple-Health-Daten für personalisierte Werbung war hypothetisch. Apple erklärt, Gesundheitsdaten aus der Health-App nicht für seine Werbeplattform zu verwenden. Bei anderen Apps und verbundenen Drittanbietern können die Regelungen abweichen. Entscheidend sind daher die jeweils geltende Datenschutzerklärung, erteilte Einwilligungen und die tatsächlich verbundenen Dienste. 01:28:28–01:28:39 Einordnung: Die Nutzung einer datenschutzkonformen Coaching-Software entbindet Coaches oder andere Behandelnde nicht automatisch von ihrer eigenen datenschutzrechtlichen Verantwortung. Je nach Ausgestaltung bleiben sie für die Verarbeitung verantwortlich und müssen unter anderem Zweck, Rechtsgrundlage, Zugriffsrechte, Auftragsverarbeitung und sichere Speicherung prüfen. Die Wahl einer geeigneten Software ist dabei nur ein Bestandteil der erforderlichen Maßnahmen. Quellenverzeichnis Alhammad, N., Alajlani, M., Abd-alrazaq, A., Epiphaniou, G., & Arvanitis, T. (2024). Patients' perspectives on the data confidentiality, privacy, and security of mHealth apps: Systematic review. Journal of Medical Internet Research, 26, Article e50715. https://doi.org/10.2196/50715 Amaral Cejas, O., Sannier, N., Abualhaija, S., Ceci, M., & Bianculli, D. (2025). GDPR-relevant privacy concerns in mobile apps research: A systematic literature review [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2411.19142 Anderberg, I., Kemps, E., & Prichard, I. (2025). The link between the use of diet and fitness monitoring apps, body image and disordered eating symptomology: A systematic review. Body Image, 52, Article 101836. https://doi.org/10.1016/j.bodyim.2024.101836 (Research @ Flinders) Aslanova, M. S., Valieva, A. S., Bogacheva, N. V., & Skupova, A. M. (2024). Mobile food tracking apps: Do they provoke disordered eating behavior? Results of a longitudinal study. Psychology in Russia: State of the Art, 17(1), 67–83. https://doi.org/10.11621/pir.2024.0104 Banal, M. G., Bongga, D., Angbengco, J. M., Amarra, S., & Panlasigui, L. (2024). MyFitnessPal smartphone application: Relative validity and intercoder reliability among dietitians in assessing energy and macronutrient intakes of selected Filipino adults with obesity. BMJ Nutrition, Prevention & Health, 7(1), 54–60. https://doi.org/10.1136/bmjnph-2023-000770 Chotwanvirat, P., Prachansuwan, A., Sridonpai, P., & Kriengsinyos, W. (2024). Advancements in using AI for dietary assessment based on food images: Scoping review. Journal of Medical Internet Research, 26, Article e51432. https://doi.org/10.2196/51432 Cofre, S., Sanchez, C., Quezada-Figueroa, G., & López-Cortés, X. A. (2025). Validity and accuracy of artificial intelligence-based dietary intake assessment methods: A systematic review. British Journal of Nutrition, 133(9), 1241–1253. https://doi.org/10.1017/S0007114525000522 Doherty, C., Baldwin, M., Keogh, A., Caulfield, B., & Argent, R. (2024). Keeping pace with wearables: A living umbrella review of systematic reviews evaluating the accuracy of consumer wearable technologies in health measurement. Sports Medicine, 54(11), 2907–2926. https://doi.org/10.1007/s40279-024-02077-2 (PubMed) Eaton, M., Probst, Y., Foster, T., Messore, J., & Robinson, L. (2024). A systematic review of observational studies exploring the relationship between health and non-weight-centric eating behaviours. Appetite, 199, Article 107361. https://doi.org/10.1016/j.appet.2024.107361 Gioia, S., Vlasac, I. M., Babazadeh, D., Fryou, N. L., Do, E., Love, J., Robbins, R., Dashti, H. S., & Lane, J. M. (2023). Mobile apps for dietary and food timing assessment: Evaluation for use in clinical research. JMIR Formative Research, 7, Article e35858. https://doi.org/10.2196/35858 Hallock, R., Ufholz, K., & Patel, N. (2024). Self-monitoring of weight as a weight loss strategy: A systematic review. Current Cardiovascular Risk Reports, 18, 163–172. https://doi.org/10.1007/s12170-024-00746-5 (DOI) Ho, D. K. N., Chiu, W.-C., Kao, J.-W., Tseng, H.-T., Lin, C.-Y., Huang, P.-H., Fang, Y.-R., Chen, K.-H., Su, T.-Y., Yang, C.-H., Yao, C.-Y., Su, H.-Y., Wei, P.-H., & Chang, J.-S. (2024). Reliability issues of mobile nutrition apps for cardiovascular disease prevention: Comparative study. JMIR mHealth and uHealth, 12, Article e54509. https://doi.org/10.2196/54509 Khazen, W., Jeanne, J.-F., Demaretz, L., Schäfer, F., & Fagherazzi, G. (2020). Rethinking the use of mobile apps for dietary assessment in medical research. Journal of Medical Internet Research, 22(6), Article e15619. https://doi.org/10.2196/15619 König, L. M., Attig, C., Franke, T., & Renner, B. (2021). Barriers to and facilitators for using nutrition apps: Systematic review and conceptual framework. JMIR mHealth and uHealth, 9(6), Article e20037. https://doi.org/10.2196/20037 McCaig, D., Elliott, M. T., Prnjak, K., Walasek, L., & Meyer, C. (2020). Engagement with MyFitnessPal in eating disorders: Qualitative insights from online forums. International Journal of Eating Disorders, 53(3), 404–411. https://doi.org/10.1002/eat.23205 Nunes, C. L., Casanova, N., Francisco, R., Bosy-Westphal, A., Hopkins, M., Sardinha, L. B., & Silva, A. M. (2022). Does adaptive thermogenesis occur after weight loss in adults? A systematic review. British Journal of Nutrition, 127(3), 451–469. https://doi.org/10.1017/S0007114521001094 (Cambridge University Press) Patel, M. L., King, A. C., Rosas, L. G., Bennett, G. G., Collins, L. M., Gallis, J. A., Zeitlin, A. B., Talreja, P. S., Crosthwaite, P. C., Collins, K. A., Lim, A. W., & Kim, T. S. (2025). Optimizing self-monitoring in a digital weight loss intervention (Spark): Protocol for a factorial randomized trial. JMIR Research Protocols, 14, Article e75629. https://doi.org/10.2196/75629 Pravst, I., Hribar, M., Žmitek, K., Blažica, B., Koroušic Seljak, B., & Kušar, A. (2022). Branded foods databases as a tool to support nutrition research and monitoring of the food supply: Insights from the Slovenian Composition and Labeling Information System. Frontiers in Nutrition, 8, Article 798576. https://doi.org/10.3389/fnut.2021.798576 Raber, M., Liao, Y., Rara, A., Schembre, S. M., Krause, K. J., Strong, L., Daniel-MacDougall, C., & Basen-Engquist, K. (2021). A systematic review of the use of dietary self-monitoring in behavioural weight loss interventions: Delivery, intensity and effectiveness. Public Health Nutrition, 24(17), 5885–5913. https://doi.org/10.1017/S136898002100358X Rezende, F., Oliveira, B. M. P. M., & Poínhos, R. (2024). Assessment of intuitive eating and mindful eating among higher education students: A systematic review. Healthcare, 12(5), Article 572. https://doi.org/10.3390/healthcare12050572 Speakman, J. R., & Hall, K. D. (2023). Models of body weight and fatness regulation. Philosophical Transactions of the Royal Society B: Biological Sciences, 378(1888), Article 20220231. https://doi.org/10.1098/rstb.2022.0231 (PubMed Central (PMC)) Stehr, P., Karnowski, V., & Rossmann, C. (2020). The multi-faceted usage patterns of nutrition apps: A survey on the appropriation of nutrition apps among German-speaking users of MyFitnessPal. BMC Medical Informatics and Decision Making, 20, Article 279. https://doi.org/10.1186/s12911-020-01294-9 Zečević, M., Mijatović, D., Kos Koklič, M., Žabkar, V., & Gidaković, P. (2021). User perspectives of diet-tracking apps: Reviews content analysis and topic modeling. Journal of Medical Internet Research, 23(4), Article e25160. https://doi.org/10.2196/25160 Zhang, L., Misir, A., Boshuizen, H., & Ocké, M. (2021). A systematic review and meta-analysis of validation studies performed on dietary record apps. Advances in Nutrition, 12(6), 2321–2332. https://doi.org/10.1093/advances/nmab058 (PubMed)

Astronomy Daily · S05E160 · 6 August 2026. In this episode: a Falcon 9 upper stage confirmed to have struck the Moon; a live ISS spacewalk; supernova SN 2026gzf caught at first light; SpaceX Crew-13 previewed; and a both-hemispheres guide to the 12 August eclipse and Perseids. FALCON 9 LUNAR IMPACT (LEAD) ●        Spent upper stage 2025-010D struck near Einstein Crater (~19.4°N, 93.3°W) at ~06:35 UTC, 5 Aug 2026 — matching Bill Gray / Project Pluto predictions. ●        ESO's Very Large Telescope (Paranal) detected sodium and lithium in the impact plume for ~5–10 minutes — sodium from lunar soil, lithium likely from the rocket. ●        NASA's Lunar Reconnaissance Orbiter holds a pre-impact baseline and will image the fresh crater, giving a rare before-and-after dataset to validate hollow-body impact models for future landers. ●        Origin: Falcon 9 launched 15 Jan 2025 with Firefly Blue Ghost 1 and ispace Resilience; the stage was left adrift and returned unplanned via solar activity and gravity. ●        Sources: NASA; ESO / VLT; CNN; Project Pluto (projectpluto.com); Fernando et al., arXiv:2607.14625. ISS SPACEWALK — U.S. EVA 96 ●        NASA's Jessica Meir (Expedition 75 commander) and Anil Menon; ~6.5 h EVA beginning ~08:35 EDT, 6 Aug; live on NASA+. ●        Task: reconfigure the 3B power channel to accept a seventh roll-out solar array later in 2026 — part of station upkeep ahead of the planned 2030 deorbit. ●        Menon's first spacewalk; two further EVAs scheduled 13 and 25 Aug. ●        Sources: NASA / NASA+ ; NASA ISS blog. SUPERNOVA SN 2026GZF ●        X-ray “shock breakout” (EP260321a) detected by China's Einstein Probe on 21 Mar 2026 from a galaxy ~500 million light-years away; follow-up revealed a broad-lined Type Ic supernova. ●        Unusually, no gamma-ray burst and no lasting relativistic jet — and the faintest shock breakout yet linked to this class; a possible missing link between ordinary and GRB-producing supernovae. ●        Follow-up used NSF NOIRLab facilities including DECam and the Vera C. Rubin Observatory. Two teams: O'Connor et al. (CMU) and Rastinejad et al. (UMD), in ApJL. ●        Sources: NSF NOIRLab; Carnegie Mellon University; The Astrophysical Journal Letters. SPACEX CREW-13 ●        Targeting NET 12 Sep 2026 aboard Crew Dragon Grace from SLC-40, Cape Canaveral. ●        Crew: Jessica Watkins (NASA, cmdr — first NASA astronaut to fly Dragon twice), Luke Delaney (NASA, pilot), Joshua Kutryk (CSA), Sergey Teteryatnikov (Roscosmos). 180-day stay; moved up from November. ●        Sources: NASA; SpaceX; Houston Public Media. SKYWATCH — 12 AUGUST & BEYOND ●        Total solar eclipse 12 Aug: totality across E Greenland, W Iceland, N Spain and NE Portugal (max 2m18s). Partial across northern US (Alaska–N Carolina), most of Canada, much of Europe and NW Africa. NASA livestream from 1:15 pm EDT. ●        US partial highlights: Presque Isle ME ~29% (contiguous-US max); NYC ~9% from ~1:07 pm EDT; Boston ~16%; Toronto ~8%; Montreal ~18%; St John's ~53%. ●        Perseids peak 12–13 Aug under a new Moon (northern-favoured, up to ~100/hr from dark northern sites). ●        Southern Hemisphere: pre-dawn planet parade (Saturn best, near opposition; Mars, Mercury, Uranus, Neptune), brilliant Venus in the evening west, and the winter Milky Way high overhead. ●        Look-ahead: deep partial lunar eclipse 28 Aug (96% umbral), favouring the Americas. ●        SAFETY: use only ISO 12312-2 certified eclipse glasses / solar filters during any partial phase. ●        Sources: NASA; ESA; Sky & Telescope; BBC Sky at Night; EarthSky; timeanddate.com; The Planetary Society.Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

S05E158 — “The Crater Makers” · Tuesday 4 August 2026. Four stories on cosmic impacts — and a new map of the X-ray sky — plus a both-hemispheres skywatch. ① A dead Falcon 9 is about to crater the Moon ●        A spent SpaceX Falcon 9 upper stage (catalogued 2025-010D) is predicted to strike the Moon near Einstein Crater on 5 Aug 2026 at ~06:35 UTC — the second documented uncontrolled rocket impact on the Moon. ●        The stage launched Firefly's Blue Ghost and ispace's Resilience landers in Jan 2025, then was stranded in a chaotic Earth orbit for ~18 months. Impact speed ~2.4 km/s (~5,400 mph); expected crater ~17–27 m across. ●        Prediction by Bill Gray (Project Pluto). NASA's LRO can image the site before and after; a ~two-dozen-author campaign is coordinating observers to chase the ejecta plume. ●        Governance angle: no international framework governs cislunar debris — a growing hazard as lunar traffic increases (callback to E154's distant-retrograde-orbit debris study). Sources: Space.com / Forbes / Reuters (Cris Tolomia) — impact preview, 31 Jul–3 Aug 2026  ·  Project Pluto (Bill Gray) — 2025-010D impact page  ·  ZME Science — impact timing, 3 Aug 2026 ② Neptune's inner moons: shrapnel of a shattered world ●        JWST NIRSpec spectroscopy of Neptune's inner moons Larissa, Galatea and Proteus (Caltech; lead author Ryleigh Davis, Mike Brown's group) detected magnesium-rich phyllosilicates — clay minerals that require liquid water and had never been seen beyond Jupiter. ●        The clays imply the material came from much larger, water-bearing worlds — evidence that Neptune's original regular moon system was destroyed when Triton (a captured Kuiper Belt object) arrived. ●        Proteus, the largest of the three, lacks the clays — possibly re-heated and reprocessed after re-forming, hiding its history. Sources: Davis et al., Science Advances, 29 July 2026  ·  Caltech / phys.org press coverage, 29 Jul–3 Aug 2026  ·  Space.com (Charles Q. Choi), 3 Aug 2026 ③ “Charbroiled within hours”: impact dust and the dinosaurs ●        New modelling (Brandon Johnson et al., Purdue) argues ultrafine impact dust from Chicxulub acted as an insulating blanket, raising surface heating ~3.5× above the spherule-only estimate — enough to ignite global wildfires and kill exposed animals within hours. ●        The same dust then blocked sunlight for years — so the model adds a ferocious opening act to the established “impact winter,” rather than replacing it. Sheltering underground or underwater aided survival. ●        Caveat: the strongest physical wildfire evidence is so far confined to North America (per UCL's Alfio Chiarenza, not involved in the study); a truly global fire record remains unconfirmed. Sources: Johnson, Johnson, Wakita & Robertson, JGR: Biogeosciences, 2026  ·  Popular Science / Space.com / ZME Science, 28 Jul–2 Aug 2026 ④ eROSITA DR2 — the X-ray sky doubles ●        The eROSITA telescope (SRG mission) released its second data set (DR2) on 31 July 2026: ~1.9 million pointlike and ~64,000 extended X-ray sources from the first three all-sky scans — nearly doubling the previously released eROSITA catalogue. ●        For scale: ROSAT catalogued ~130,000 sources in the 1990s and was the benchmark for 30 years. Lead author: Miriam Ramos-Ceja (MPE). ●        DR2 delivers the first direct census of cataclysmic-variable binaries sufficient to explain the decades-old Galactic Ridge X-ray Emission. It coincides with SDSS DR20 (Black Hole Mapper), enabling 3D mapping of active black holes. Sources: Ramos-Ceja et al. (eROSITA-DE), arXiv:2607.27772; DR2 release  ·  Max Planck Society press release, 29 July 2026  ·  phys.org, 31 Jul 2026 Skywatch — both hemispheres ●        Falcon 9 impact (5 Aug, ~06:35 UTC): not naked-eye; ejecta plume a long shot for advanced amateurs in the Americas near the western limb (~2:35 a.m. EDT). Sydney: Moon below horizon — watch for LRO after-images. ●        Comet 10P/Tempel 2: perihelion 2 Aug, closest to Earth 3 Aug (~0.41 AU); ~mag 8–9 near M30 in Capricornus. Best return of the century; Southern-Hemisphere-favoured; Northern viewers low on the southern horizon. Darkest window ~7–8 Aug. ●        This evening: Venus threads between Regulus (Leo) and Spica (Virgo) in the west — both hemispheres, no equipment. Mercury at its best morning showing (greatest western elongation 2 Aug), low in the pre-dawn east. ●        Diary — 12 Aug: total solar eclipse (Greenland/Iceland/Spain), moonless Perseid peak, six-planet alignment. Always use certified ISO 12312-2 eclipse glasses / solar filters except during totality.Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

We explore the state of education just before the rise of AI, compare it to today, and extrapolate into the future. SourcesTyler Jagt, "My Students Can't Read," Chronicle of Higher Education, 2026. NAEP Long-Term Trend AssessmentPew Research, "Reading for fun has become less common," 2021.James S. Coleman et al., Equality of Educational Opportunity ("The Coleman Report"), US Dept. of Health, Education, and Welfare, 1966Kathryn Borman & Monique Dowling, Teachers College Record, 2010.Dan Goldhaber, "The Mystery of Good Teaching," Education Next, 2002 Steven Rivkin, Eric Hanushek & John Kain, teacher-effects study, Econometrica, 2005. Xitao Fan & Michael Chen, parental-involvement meta-analysis, Educational Psychology Review, 2001 World Bank, "Vietnam's Human Capital: Education Success and Future Challenges," 2020 (PISA science outperformance on low per-pupil spend). OECD, "Does Money Buy Strong Performance in PISA?", PISA in Focus No. 13, 2012 (no score/spending relationship above ~$35k cumulative per-student spend). NCES, Digest of Education Statistics 2023, Table 236.55 (US per-pupil spending, $6,873 → $17,495 real dollars, 1970-71 to 2020-21)Boston Globe, 2001 coverage of Harvard grading ("laughing stock of the Ivy League"; 91% of seniors graduating with honors). Harvard Crimson, Oct. 5, 2023 (~79-80% of 2020-21 grades in the A range)Anemona Hartocollis, "Harvard Students Skip Class and Still Get High Grades, Faculty Say," New York Times, Oct. 6, 2025. Gallup, confidence-in-higher-education tracking, July 2023 and July 2024 releases (57% in 2015 → 36% in 2023/24; Republican confidence ~56%→~19-20%, Democratic decline much smaller)Pew Research, Aug. 2019 (share saying colleges have a negative effect on the country, 26%→38%, 2012-2019, driven almost entirely by Republicans/leaners)Challenge Success (Stanford, Denise Pope), Fall 2018 surveyMark Galloway, cheating in advantaged-community schools, Ethics & Behavior, 2012Stacy Dale & Alan Krueger, "Estimating the Payoff to Attending a More Selective College," NBER WP 7322, Aug. 1999; Quarterly Journal of Economics, 2002. Follow-up: NBER WP 17159, Jun. 2011; Journal of Human Resources, 2014.ProPublica & The Chronicle of Higher Education, "The Newest College Admissions Ploy: Paying to Make Your Teen a 'Peer-Reviewed' Author," May 2023.Nataliya Kosmyna et al., "Your Brain on ChatGPT," MIT Media Lab, arXiv:2506.08872, 2025. Hao-Ping (Hank) Lee et al., "The Impact of Generative AI on Critical Thinking," CHI 2025, peer-reviewed, n=319. Turnitin AI-detection data, 2023-2026 (>80% AI-written flags: ~3% early 2023 → ~15% early 2026).Inside Higher Ed / Generation Lab Student Voice survey, Aug. 2025Rose Horowitch, "The Elite College Students Who Can't Read Books," The Atlantic, Oct. 2024. "Hilarius Bookbinder," "The Average College Student Is Illiterate," Persuasion, Mar. 2025. Tyler Jagt, "My Students Can't Read," Chronicle of Higher Education, 2026. NAEP 2024 12th-grade reading assessment, released Sept. 2025 (lowest since assessment began). Erik Brynjolfsson, Bharat Chandar & Ruyu Chen (Stanford Digital Economy Lab), "Canaries in the Coal Mine?," Nov. 2025, ADP payroll dataIndeed Hiring Lab, entry-level software postings2U, coding-bootcamp partnership shutdown, Dec. 2024.Southern New Hampshire University, bootcamp closure, 2023, citing AI adoption.World Economic Forum, Future of Jobs Report, 2023 & 2025 McKinsey Global Institute, "Skill Shift," 2018 Stefan Bauschard, "Education is a Wicked Problem in AI" Brian Miller, "Farewell, Peak Literacy, We Hardly Knew You," Front Porch Republic, Nov. 2023. James Marriott, The New Dark Ages: The Death of Reading and the Dawn of the Post-Literate Society Hosted on Acast. See acast.com/privacy for more information.

Astronomy Daily · S05E157 · “The Edge of Everything” · Monday 3 August 2026 In this episode •      The largest galaxy, measured at last. Ultra-deep imaging of IC 1101 in the Abell 2029 cluster has traced its outer edge for the first time: an edge radius of ~260 kpc, a diameter near 520 kpc (~1.7 million light-years), and ~3.4 trillion solar masses in stars — about 17× the width of the Milky Way. Faint outer structures aligned with cluster X-ray disturbances show it is still accreting. Source: Marrero-de la Rosa et al., “How large can galaxies be? Ultra-deep imaging of IC 1101”, accepted A&A (arXiv:2607.15340, 16 Jul 2026); phys.org, 31 Jul 2026; The Debrief, 1 Aug 2026. •      Hayabusa2's record asteroid flyby. JAXA confirmed (press conference 30 Jul) that its Hayabusa2 probe passed just 400 m from the surface of near-Earth asteroid Torifune on 5 July — 744 m from centre at >18,000 km/h — the closest asteroid flyby by any mission, plus the first-ever laser ranging of an asteroid during a flyby. Torifune proved to be a contact binary. Source: JAXA press conference, 30 Jul 2026; phys.org / Japan Today (AFP), 31 Jul 2026; autoevolution, 1 Aug 2026. •      UK's first orbital launch on hold. Rocket Factory Augsburg de-stacked its RFA One at SaxaVord Spaceport (Unst, Shetland) after finding a vehicle issue during hot-fire test prep (announced 28 Jul). The certified launch window opens 10 August; a successful flight would be the first orbital launch from UK soil. Source: RFA statement (X), 28 Jul 2026; Space.com, 1 Aug 2026; Shetland Times, 28 Jul 2026. •      Andromeda is winding down. A new Hubble study mapping ~200 million stars across two-thirds of Andromeda's disk finds star formation has declined over ~500 million years (from ~1 to ~0.2 solar masses/year), with the steepest drop in the last 40 million years — possibly linked to an interaction with satellite galaxy M32. Source: Williams et al., The Astrophysical Journal, 27 Jul 2026; NASA/STScI release, 27 Jul 2026. Skywatch — both hemispheres •      Comet 10P/Tempel 2 — perihelion 2 Aug; closest approach to Earth tonight (3 Aug). Binoculars/small scope, dark skies. Best pre-dawn from the Southern Hemisphere; ~9:30–10:30 pm local for mid-northern latitudes before moonrise. •      Moon near Saturn tonight (3 Aug) — naked-eye pairing, both hemispheres, up into dawn. •      Mercury at greatest western elongation (2 Aug) — low eastern horizon ~1 hr before sunrise, toward Gemini; Mars higher. Better placed for northern observers. •      Diary (2 days out): Falcon 9 upper stage (2025-010D) lunar impact near Einstein Crater, 5 Aug ~06:35 UTC — ~2:35 am EDT / 11:35 pm PDT (4 Aug); North America best-timed on the dark limb. Southern Hemisphere: ~4:35 pm AEST (daylight) — await orbiter after-images. •      Looking ahead: 12 Aug total solar eclipse (Greenland/Iceland/Spain) + moonless Perseid peak + six-planet dawn alignment. View partial phases only through certified ISO 12312-2 filters. Links & follow •      Website, back catalogue, news feed & newsletter: astronomydaily.io •      Social: @AstroDailyPod  ·  Part of the Bitesz.com Podcast NetworkBecome a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Het internet is kwetsbaarder dan je misschien denkt; één programmeur liet al een keer een deel van het internet crashen, waar giganten als Netflix en Spotify last van hadden. In deze aflevering laat softwareonderzoeker Ayushi Rastogi (Rijksuniversiteit Groningen) je de wereld zien achter de apps en websites die jij elke dag gebruikt. Daar wordt namelijk wereldwijd aan gebouwd en gaat niet alleen over techniek, maar vooral ook over menselijk vertrouwen. Dat maakt het internet krachtig, maar is ook haar zwakte. Ayushi onderzoekt hoe we dat internet in de toekomst betrouwbaar en eerlijk voor iedereen houden. Want in een tijd waarin AI razendsnel nieuwe code genereert, maar ook onvoorspelbare fouten maakt, is dat één van de belangrijke vragen van deze tijd. Omdat veel wetenschappers die in Nederland werken Engels spreken, maar we hun onderzoek ook willen laten zien, hebben we voor het eerst in lange tijd een aflevering die deels in het Engels is. Voor wie toch liever in het Nederlands luistert, is dit de link naar de nagesynchroniseerde versie op YouTube: https://youtu.be/G1W-at8UFns. 0:00 Kan één iemand het hele internet laten crashen?0:54 Wat is het internet?1:19 Hoe werkt programmeertaal?1:46 Wie werkt er aan het internet?1:53 Hoe werkt open source software?2:26 Wat gebeurt er als een programmeur bepaalde codes weghaalt?3:48 Kunnen programmeurs niet elke keer zelf hun code schrijven?4:38 Wat zijn gate keepers in programmeren?5:16 Wat gebeurde er tijdens de laatste grote wereldwijde IT-storing?6:01 Wat is het Blue Screen of Death (BSoD)?6:10 Waarom maakt het uit wie software ontwikkelt?7:15 Waarom is er ongelijkheid tussen mannelijke en vrouwelijke ontwikkelaars?8:08 Wat gebeurt er als AI onze codes gaat schrijven? Benieuwd naar de studies die Ayushi noemt? Je vindt ze hier: Article on how self-tracking apps exclude women https://www.theatlantic.com/technology/archive/2014/12/how-self-tracking-apps-exclude-women/383673/ Study showing how factors influencing the decision to accept or reject incoming software changes are both technical and social. https://doi.org/10.1109/TSE.2022.3165056 and https://arxiv.org/pdf/2106.01885 Study showing how contributions from women are evaluated in comparison to their male counterparts: https://doi.org/10.7717/peerj-cs.111 Similar differences were seen across geographical location https://doi.org/10.1145/3239235.3240504 Humans are likely to miss errors AI introduces: https://doi.org/10.48550/arXiv.2607.08885 and https://doi.org/10.1109/CHASE66643.2025.00016 Future of human-AI collaboration: https://doi.org/10.1016/j.jss.2025.112699 Balancing Speed, Reliability, and Fairness: https://doi.org/10.1145/3786583.3786888 and https://doi.org/10.1145/3757060 and https://doi.org/10.1007/s10664-023-10401-z See omnystudio.com/listener for privacy information.

S05E155 · Friday 31 July 2026 · Astronomy Daily with Anna & Avery. Four stories and a both-hemispheres skywatch. Australian English. ① Asteroid (44) Nysa — the first three-lobed world, and its hidden moon ●        trilobate body — three lobes joined by two narrow necks; a candidate “first” of its kind. ●        A new ~1 km moon, S/2026 (44) 1, was found orbiting ≥170 km out — spotted using high-contrast imaging borrowed from exoplanet work, and confirmed moving across two observing runs. ●        The moon lets astronomers weigh Nysa (mass → density), which should help decide between a genuine contact-trinary and a single, deeply indented body. ●        Nysa is a bright, main-belt E-type (enstatite-rich) asteroid, ~75 km across, known since 1857. ●        Source: Lowell Observatory / University of Arizona press release, 29 Jul 2026; study “Unmasking (44) Nysa: Evidence for a Trilobate Structure” (Minker et al.). Coverage: Space.com, Gizmodo, 29–30 Jul 2026. ② Mapping Alien Continents — a NASA concept to image an exoplanet's surface ●        NASA's 2026 NIAC round funds 18 early-stage “visionary” concepts (~$175k each, 9 months). These are seed studies, not missions. ●        Paul Stankus (Brookhaven) proposes “Mapping Alien Continents”: resolve the surface of an Earth-like exoplanet — continents, oceans — in visible light. ●        Method: a novel “dynamic hierarchical nulling” interferometer to suppress starlight at 10¹⁰-to-1 contrast, then combine beams from two spacecraft ~100 km apart (optical VLBI-style imaging). ●        Source: NASA “2026 Innovative Technology Concepts” release and NIAC selections (posted 21 Jul; consolidated release ~29 Jul 2026); Universe Today feature, 30 Jul 2026. ③ Solar-storm watch — CMEs inbound, minor-storm and aurora potential ●        Two faint coronal mass ejections, plus coronal-hole solar wind, are set to give Earth glancing blows; forecasters flag possible G1 (minor) geomagnetic storms and auroras over the coming days. ●        G1 means little grid impact but aurora visible at somewhat lower latitudes than usual — see the skywatch for where to look, both hemispheres. ●        Source: NOAA SWPC (WSA-ENLIL model); EarthSky / The Sun Today, 30 Jul 2026. ④ ESCAPADE's family portrait of Earth and the Moon ●        NASA's twin Mars orbiters (“Blue” and “Gold,” built by Rocket Lab) imaged Earth and the Moon as thin crescents in visible and thermal-infrared light from a loiter orbit near Sun–Earth L2. ●        In infrared, Earth's night side glows with its own heat; the Moon's shadowed half is far colder — a calibration check before Mars. ●        ESCAPADE's science goal (arrival Sept 2027): measure how the solar wind strips Mars's unshielded atmosphere — the payoff of today's solar-wind thread. ●        Source: NASA (Goddard) image feature, ~25 Jul 2026 (images captured 3 Jul); NAU / phys.org; Universe Today, 29 Jul 2026. Skywatch — both hemispheres ●        Bright waning-gibbous Moon (post-Buck-Moon, 29 Jul) washes out faint targets. ●        Southern Delta Aquariids + Alpha Capricornids just past peak (SH-favoured; Moon-hampered). Perseids build to a near-moonless peak on the night of 12–13 Aug — prime for North America, low in the north for the SH. ●        12 Aug total solar eclipse: totality across Greenland / Iceland / Spain; partial for parts of northern North America and Europe. ISO 12312-2 eye protection required for any partial phase. ●        Pre-dawn planets low in the east (Sydney and North American framing). Aurora watch for high latitudes both hemispheres if the storms land. ●        Diary: a spent Falcon 9 upper stage (2025-010D) is predicted to hit the Moon near Einstein Crater on 5 Aug 2026, ~06:35 UTC (~2:34 a.m. ET). North America best-timed; telescope needed, Moon ~56% lit, target the limb dust plume. From Sydney the Moon is down at impact — rely on LRO after-images. Callback to E125/E147. Source: Fernando et al., arXiv 2607.14625.Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Astronomy Daily S05E154 — “The Reach of a Giant.” Thursday, 30 July 2026. Hosted by Anna and Avery. In this episode •    A supermassive black hole caught heating and stirring an entire galaxy cluster — turbulence reaching ~300,000 light-years, about 100× more energetic than expected (XRISM / quasar H1821+643). •    Why some “clean” white dwarfs are secretly feasting: magnetic fields funnel shredded planetary debris to the poles, aurora-style, hiding the meal. •    NASA's Psyche used its Mars flyby as a dress rehearsal — spotting Phobos and Deimos from afar to practise for its 2029 moonlet hunt at asteroid 16 Psyche. •    A new study on keeping the road to the Moon clear: modelling how debris disperses in Distant Retrograde Orbits as cislunar traffic climbs. •    Skywatch: tonight's double meteor shower — the Southern Delta Aquariids and Alpha Capricornids — with both-hemispheres viewing details and local times. Sources •    Yamada, S. et al., “Vigorous turbulence driven by quasar-mode feedback in a cluster core,” Nature Astronomy, 28 July 2026 (DOI 10.1038/s41550-026-02939-x; arXiv 2607.24911). Tohoku University release. •    “White Dwarfs Eat More Planetary Debris Than Thought, But Magnetic Fields Hide It,” Universe Today, 29 July 2026 (Pham et al., arXiv 2607.20747). •    “NASA's Psyche Spacecraft Aces Mars Flyby,” NASA JPL / ScienceDaily, 28 July 2026. •    “The Risks of Debris Between the Earth and the Moon for Future Exploration,” Universe Today, 29 July 2026 (Chinese Academy of Sciences DRO study). •    Double meteor shower peak: American Meteor Society; NASA; Scientific American; CNN; National Geographic, 28–30 July 2026. Correction / caveat desk •    Skywatch numbers assume the ~98% waning Buck Moon (full 29 Jul). Faint Delta Aquariids will be washed out; the Alpha Capricornid fireballs are the reliable catch tonight.Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Journée Grangier le 4 juin à l'IOGS Journée dédiée à Philippe Grangier qui retraçait le parcours scientifique de ce dernier avec des témoignages des scientifiques qu'il a encadrés (thésards, post-docs) et avec qui il a collaboré. Programme https://grangier26.sciencesconf.org/resource/page/id/1Vidéos https://www.youtube.com/live/7aQScUtzvdE?si=ix_s-t0HppbHqw0tSlides Olivier https://www.oezratty.net/Files/Conferences/Olivier%20Ezratty%20AI%20for%20Philippe%20Grangier%20Day%20Jun2026.pdf France Quantum le 16 juin à Station F Un beau succès avec 1200 participants et plus de 20 stands.Prix du chercheur à Julien Laurat et de la startup de l'année à Quobly.Replay https://www.youtube.com/watch?v=005TX0IPqMA&list=PLHy9A3t7TeETvyZJIvPkexfFmwVjChkkeSlides Olivier https://www.oezratty.net/Files/Conferences/Olivier%20Ezratty%20France%20Quantum%20FTQC%20challenges%20June2026.pdf Vivatech du 16 au 20 juinReplay table ronde cybersécurité https://www.dailymotion.com/video/xagml3m IQT Nordics à Oslo des 22 au 24 juinSlides Olivier https://www.oezratty.net/Files/Conferences/Olivier%20Ezratty%20QEI%20IQT%20Nordics%20Jun2026.pdf Panorama de toutes les voies technologiques de l'ordinateur quantique à Grenoble les 25 et 26 juinSéminaire d'une journée et demie organisée par la Maison du Quantique Grenoble-Alpes / le Hub Quantiquehttps://quantalps.univ-grenoble-alpes.fr/quantum-technologies-deep-dive-1713265.kjsp Quantum Koreahttps://quantum-korea.kr/ko/main ContenusCours de Jean Dalibard au Collège de France https://www.college-de-france.fr/fr/agenda/cours/fluides-quantiques-couples-et-jonctions-josephsonUn podcast d'Alain Aspect avec l'Académie Française. https://podcasts.institutdefrance.fr/emissions/en-habit-vert/de-la-physique-quantique-a-lacademie-francaise-avec-alain-aspect-prix-nobel-de-physique Evénements à venir IEEE Quantum Week à Toronto des 13 au 18 septembre (agenda). Quantum Datacenter Alliance forum 2026, October 1st, Londres. J'y interviens pour la QEI.QEC 2027 sera à Paris, des 7 au 11 juin 2027. Les soumissions de contributions doivent se faire d'ici le 31 janvier 2027.Journée QuantAlps à Grenoble, 8 et 9 octobre (inscription).Forum Teratec 20 et 21 octobre au Palais des Congrès à Paris.GDR TEQ 2026 à l'ENS Paris Saclay les 4, 5 et 6 novembre (inscriptions). 110€ pour les gens normaux, 55€ pour les étudiants.World Quantum Cannes Festival, 17 et 18 novembre (agenda).EQTC 2026 à Dublin du 30 novembre au 3 décembre. J'y interviens au titre du Q-CAB. France Levée de fonds de Quobly$115M. Troisième startup française à lever plus de 100M€ après Pasqal et Alice&Bob. https://presse.bpifrance.fr/quobly-leve-115-millions-deuros-en-serie-a-pour-commercialiser-ses-ordinateurs-quantiques-sur-silicium/ Alice & Bob Ils communiquent sur la notion de qubits logiques. Mais attention à ne pas oublier les fonctionnalités de la FTQC.https://alice-bob.com/blog/not-all-logical-qubits-are-created-equal-heres-how-to-benchmark-them/https://alice-bob.com/wp-content/uploads/2026/06/Five-criteria-to-benchmark-logical-qubit-claims.pdfHelium est officiellement lancé avec 18 qubits physiques, pour l'instant qui fonctionnent chez Alice & Bob et doivent être livrés au TGCC en 2027.https://alice-bob.com/blog/this-is-not-a-product-launch/ Papier de Xavier Waintal sur le quantum annealingNotion de bruit et d'erreur intégrée dans l'appréciation.A fidelity metric for quantum annealing benchmarked by extreme scaling quantum Monte-Carlo simulations by Gabriel Gouraud, Miha Srdinsek, and Xavier Waintal, arXiv, June 2026 (19 pages). Rapport Cour des Comptes sur le quantiqueLe rapport fait 128 pages. https://www.ccomptes.fr/fr/publications/le-soutien-linformatique-quantique-premier-bilan-et-perspectivesPapier de Victor Kerros dans Quantum Times sur la consumérisation du calcul quantiquehttps://quantum-times.com/will-quantum-computing-ever-become-consumer-technology/ International EuropeFanny Bouton et Elham Kashefi intègrent le Q-CAB. Le draft de la roadmap EU SRIA 2026 est disponible pour commentaires. https://qt.eu/about-quantum-flagship/SRIA USADeux executive orders de POTUS publiés le 22 juin. Un sur la PQC et un autre sur l'effort dans le calcul quantique dont le niveau d'intérêt stratégique augmente.

Astronomy Daily — S05E147: "Space Mechanic" Wednesday 22 July 2026 A spacecraft with robotic arms is on its way to geostationary orbit to keep other satellites alive. A discarded rocket stage is two weeks out from hitting the Moon, and twenty-three astronomers have just asked the world to watch. Plus the first binary star system where both stars exploded, the first complete magnetic map of a galaxy cluster, and the asteroid breakup that may have bombarded three worlds while Earth froze. In This Episode ●      The Space Mechanic Launches — Northrop Grumman's Mission Robotic Vehicle lifted off from Cape Canaveral on 21 July carrying three Mission Extension Pods. With two 3-metre robotic arms built by the US Naval Research Laboratory, it is designed to inspect, relocate, repair and refuel satellites in geostationary orbit. Each pod can give a 2,000 kg satellite up to eight more years of life. ●      UPDATE — The Rocket Aimed at the Moon — A new arXiv preprint signed by 23 astronomers calls for a coordinated observing campaign when Falcon 9 upper stage 2025-010D strikes the Moon near Einstein crater on 5 August. North America has the best seat: 2:34am EDT, with the paper naming observers in the Americas as the ideal group. Refined impact time, predicted crater size, and why the ejecta plume may be visible even if the flash is not. ●      Sibling Supernovae — Sixteen years of Fermi data reveal a faint supernova remnant hiding in the glare of the Jellyfish Nebula. The two may be the first known pair of remnants traced back to a single binary star system. ●      Mapping a Cluster's Magnetic Field — Using the deepest radio observations ever made with LOFAR, astronomers have reconstructed the magnetic field of galaxy cluster Abell 2255 from nucleus to outer edge for the first time. ●      The Eulalia Bombardment — A new Planetary Science Journal paper links the breakup of a main-belt asteroid to an impact shower that battered the Moon, Earth and Mars 800 million years ago — and may connect to a global freeze. ●      Skywatch, Both Hemispheres — Why this week beats peak night for the Delta Aquariids wherever you are, how to catch them from the southern US and Mediterranean, and what's coming on 12 August: a total solar eclipse across Iceland and Spain, a North American partial, and the best Perseid peak in years on a new Moon.Sources & Further Reading ●      Space.com — SpaceX launches satellite repair drone with 10-foot robotic arms to Earth orbit ●      NASASpaceflight.com — Falcon 9 to launch MRV-1 robotic servicing spacecraft for Northrop Grumman ●      Northrop Grumman SpaceLogistics — Mission Robotic Vehicle and Mission Extension Pod fact sheets ●      Scientific American — A SpaceX rocket is about to crash into the moon; scientists will be watching ●      Phys.org — When a SpaceX rocket crashes into the moon, scientists will be watching (arXiv preprint) ●      Project Pluto (Bill Gray) — Upper stage impacting the moon on 2026 August 5 ●      Stanford University — Researchers uncover evidence for sibling supernovas (Michailidis et al., Nature Communications) ●      Reuters — Scientists spot evidence of two huge companion stars that blew up ●      Space.com — Galaxy cluster's magnetic field reconstructed for 1st time with record-breaking astronomy map (Botteon et al., INAF, A&A) ●      Southwest Research Institute — SwRI-led research connects asteroid collision to impact showers 800 million years ago ●      The Planetary Science Journal — Bottke, Vokrouhlický, Dykhuis & Zellner, "An 800 Myr-old Impact Shower on the Terrestrial Planets from the Breakup of the Eulalia Parent Body" ●      EarthSky — Delta Aquariid meteor shower: all you need to know in 2026 ●      NASA Science — Total Solar Eclipse on August 12, 2026 (path, partial visibility and safety guidance) ●      BBC Sky at Night Magazine — August 12, 2026 solar eclipse: USA and Canada guide   Connect ●      Website: astronomydaily.io ●      Socials: @AstroDailyPod ●      Part of the Bitesz.com Podcast NetworkBecome a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.astronomydaily.io/nordvpn. You'll be glad you did!Get the best secure and private email on the planet. Stop your Government, google and who knows who else spying on every email you write. Do what we did and use ProtonMail. They beleive in privacy and there are no ads in their business model...yet they still provide a free forever service. Check them out and get out special deal at www.astronomydaily.io/protonmailBecome a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

We're talking about that bad article on critical confabulations in archives. We talk about history as a process, the archive, gaps in our knowledge, language as a conveyor of reality, and The Watermelon Woman.  Media mentioned THE ARTICLE: Critical Confabulation: Can LLMs Hallucinate for Social Good? https://arxiv.org/abs/2511.07722 Guillaume Cabanac, “How a Tortured Conference Becomes a Series: An Analysis of Conference Manipulations” IEEE Xplore  https://ieeexplore.ieee.org/document/11363742  Research repository ArXiv will ban authors for a year if they let AI do all the work, TechCrunch  https://techcrunch.com/2026/05/16/research-repository-arxiv-will-ban-authors-for-a-year-if-they-let-ai-do-all-the-work/  The Watermelon Woman https://en.wikipedia.org/wiki/The_Watermelon_Woman  Keith Haring Unfinished Painting: https://mymodernmet.com/artificial-intelligence-finishes-keith-harings-unfinished-painting/ Patrick Wyman, Lost Worlds https://www.harpercollins.com/products/lost-worlds-patrick-wyman?variant=43084775817250  Predatory States: Operation Condor and Covert War in Latin America, Patrice McSherry https://www.bloomsbury.com/us/predatory-states-9780742568709/  Dorothea series on patron privacy https://ischool.wisc.edu/continuing-education/tech-crash-course-data-information-privacy/  Transcript: https://pastecode.io/s/3oy0sgqw  Join the Discord: https://discord.gg/qWPTurTnkT

After a short break due to busy calendars, we are back to discuss the role automation and artificial intelligence can play in science. Is AI evil, or does it just amplify the problems that were already present in science? Can automated AI driven checks help scientists to improve the way they work? If so, where are these tools helpful, and when? Enjoy.  The Babbage quote is from Babbage in November 1839, recalling events in 1821; quoted in Harry Wilmot Buxton and Anthony Hyman (1988), Memoir of the Life and Labours of the Late Charles Babbage.   Alfred Whitehead: An Introduction to Mathematics (1911)   Blog about AI references: Evaluating Dr. Cuddy's Claim that the Debunking of Power Posing is a Myth. https://daniellakens.blogspot.com/2026/05/evaluating-dr-cuddys-claim-that.html   Metacheck: DeBruine L, Mesquida C, Werner J, Lakens D (2026). metacheck: Check Research Outputs for Best Practices. doi:10.5281/zenodo.20704754, R package version 0.1.0, https://scienceverse.org/metacheck   Cummins, J., Clarke, B., Hussey, I., & Elson, M. (2026). RegCheck: A tool for structured comparisons between study registrations and papers (arXiv:2601.13330). arXiv. https://doi.org/10.48550/arXiv.2601.13330  

Exact Sciences are things folks who consider themselves rational and logical assume can be trusted. And they would be, but scientists are only human, and some of them aren't even autistic, gasp! I had the privilege of pursuing an academic career in Natural Sciences, and there I found hard-working, dedicated professors, who would base their calculations on values with dubious origins, and reject valid findings to scoop the author's work and publish it as their own. In this scientific episode I will be joined by Dr. Mary Sims (pronouns: she/her), and together we will spill the tea about science shenanigans, and on the way, you might even learn some random details about atmospheric aerosols, vibrational spectroscopy and self-assambled monolayers. You can find Dr Mary Sims' podcast, Clinical Misfits on the Autistic Culture Podcast Network here.You can support Aut2Aut on Betterplace and Gofundme, or buy our #ActuallyAutistic designs in our print-on-demand shop. This will help prepped.to go on providing a platform for autistic folks to share locations' sensory info and service instructions.My cat Merlin has a message for you at the end of the episode.Mentioned in this episode:Dr. Mary Sims' podcast, Clinical Misfits [APCN please insert link here]My paper about the refractive index of Elemental CarbonMy paper about using the Vibrational Stark effect to investigate Self-Assambled MonolayersFafarman et al. 2010 about the effect of hydrogen bonding on the Star tuning rate. Correction: the numbers I'm referring to are not in a table but a chart in their supplementary material section.Where physicists upload their papers without peer review: arXiv.orgAut2Aut, the nonprofit I founded, providing free platforms and content by and for autistic peopleprepped.to is the website I created where autistic folks can upload and consult sensory info and service instructions about places, so folks can prep and script before going thereHow to support my nonprofitTheme music composed and produced by Lir Lutau Shahar (pronouns: he/fae/sea): YouTube, Soundcloud. For collaborations: lutaoshzh@gmail.com Follow Dr. Gal Schkolnik on LinkedIn, Mastodon or Tumblr

In this episode of Astronomy Daily, Anna and Avery explore six remarkable stories from the frontiers of space science. JWST has determined that interstellar comet 3I/ATLAS likely formed 10–12 billion years ago — before our Sun existed — making it the oldest object ever chemically characterised. NASA's Perseverance rover has delivered its most robust organic detection yet in Mars's Jezero Crater. ESA's Euclid telescope has released the largest and most detailed visible-light image ever taken of the Milky Way's galactic bulge. NASA's plan to deorbit the ISS into the Pacific Ocean faces new legal and environmental scrutiny. Research from the University of Glasgow reveals the Chicxulub impact crater hosted an underground hydrothermal system for eight million years — four times longer than previously estimated. And astronomers have discovered the first-ever pair of sibling supernova remnants, hiding in the glow of the famous Jellyfish Nebula.   Story 1 — JWST & 3I/ATLAS Origin • Cordiner et al. (2026). 'Isotopic evidence for a cold and distant origin of 3I/ATLAS.' Nature. DOI: 10.1038/s41586-026-10771-6 • Opitom et al. (2026). 'High nitrogen and carbon isotopic ratios in the interstellar comet 3I/ATLAS.' Nature (in press). arXiv: 2603.07187 • NASA Science: https://science.nasa.gov/missions/webb/nasas-webb-finds-clues-to-ancient-distant-origin-of-comet-3i-atlas/ • Science Magazine: https://www.science.org/content/article/interstellar-comet-unlike-anything-seen-our-solar-system   Story 2 — Perseverance Organic Detection • Murphy et al. (2026). 'Spatially distributed complex organic matter detected in an ancient river valley in Jezero crater, Mars.' Science Advances. DOI: 10.1126/sciadv.adx0047 • Space.com: https://www.space.com/astronomy/mars/did-nasa-just-find-evidence-of-ancient-life-on-mars-perseverance-rover-spots-complex-carbon-in-red-planet-rocks • ScienceAlert: https://www.sciencealert.com/perseverance-finds-complex-organic-compounds-in-strange-mars-rocks   Story 3 — Euclid Milky Way Image • ESA Euclid Mission Press Release, 24 June 2026 • NASA JPL: https://www.nasa.gov/missions/roman-space-telescope/euclid-view-of-milky-way-heart-previews-core-survey-by-nasas-roman/ • Space.com: https://www.space.com/astronomy/galaxies/this-is-the-largest-and-most-detailed-image-of-our-milky-way-with-over-60-million-stars-and-50-exoplanet-systems • CBS News: https://www.cbsnews.com/news/euclid-telescope-most-detailed-image-milky-way-stars/   Story 4 — ISS Deorbit Environmental Concerns • US Government Accountability Office report on ISS deorbit, June 2026 • Space.com: https://www.space.com/space-exploration/international-space-station/nasa-wants-to-dump-the-iss-in-the-sea-experts-say-the-plan-raises-serious-concerns-for-ocean-health • The Ocean Foundation statement, June 2026   Story 5 — Chicxulub Hydrothermal System • Pickersgill et al. (2026). 'Hydrothermal activity persisted for at least 8 Myr at Chicxulub.' Communications Earth & Environment. DOI: 10.1038/s43247-026-03618-5 • Phys.org: https://phys.org/news/2026-06-dino-asteroid-fueled-underground-life.html • EarthSky: https://earthsky.org/earth/dinosaur-killing-asteroid-underground-hydrothermal-habitat/   Story 6 — Jellyfish Nebula Sibling Remnant • Astrophysicists' paper on IC 443 sibling supernova remnant, Universe Today, June 23 2026 • Universe Today: https://www.universetoday.com/Become a supporter of this podcast: https://www.spreaker.com/podcast/astronomy-daily-the-latest-space-news--5648921/support.Sponsor Details:Ensure your online privacy by using NordVPN. To get our special listener deal and save a lot of money, visit www.bitesz.com/nordvpn. You'll be glad you did!Become a supporter of Astronomy Daily by joining our Supporters Club. Commercial free episodes daily are only a click way... Click HereThis episode includes AI-generated content.

Panini-Sticker sind für viele Kindheit, WM und Glitzer. Aber auch Zufall, Frust und Marktlogik. Gregor und Bo gehen der Frage nach: Wie bringt uns Panini zum Weiterkaufen und war das unser erster Kontakt mit Wirtschaft?**********Zusätzliche InformationenHörtipp: Unboxing News “WM 2026 – Wie Gianni Infantino die FIFA beherrscht**********In dieser Folge:1:57 - Paninis Geschichte und Geschäftsmodell6:07 - Mit wie viel Geld kriegt man das Album voll?13:41 - Paninis Zukunft: Neue Konkurrenz im Stickerbusiness18:50 - Fazit / Wahres für Bares20:27 - Hörtipp: Unboxing News**********An dieser Folge waren beteiligt: Autor*innen: Gregor Lischka und Bo Hyun Kim  Redaktion: Michael Böddeker  Produktion: Choukri Gustmann **********Die Quellen zur Folge:Martinez, E. Z., Achcar, J. A., Peres, M. V. de O., Oliveira, R. P. de & Pose, R. A. (2022). How much does it cost to complete a sticker collection? A simulation study using R. arXiv.Anzolin, E. (2026). Italy's Panini owners to decide by year-end on possible sale, source says. Reuters, 2. April 2026.**********Unsere Empfehlungen:Peter Alexander & Deutsche Nationalelf - Mexico Mi Amor (Wetten Dass..?) 12.04.1986  **********Weitere Beiträge zum Thema:Rekord-Börsengang: Elon Musk, SpaceX und die Raumfahrt-MilliardenSanktion per Klick: Wenn die USA den Stecker ziehen Roboter-Revolution: Per Kung-Fu Kick in die Arbeitslosigkeit?**********Habt ihr auch manchmal einen WTF-Moment, wenn es um Wirtschaft und Finanzen geht? Wir freuen uns über eure Themenvorschläge und Feedback an whatthewirtschaft@deutschlandfunknova.de.**********Den Artikel zum Stück findet ihr hier.**********Ihr könnt uns auch auf diesen Kanälen folgen: TikTok und Instagram .

I keep being accused of using AI. I've even been accused, more than once, of being AI-generated. So I owe you something better than irritation: an actual explanation of where I stand.In this episode, I work through the real concerns: the scraping of artists' work, the environmental cost, algorithmic bias, the fear of job displacement, the worry about deskilling, and argue that every one of them is a problem of how, not of whether. They are arguments for regulation, not for personal abstention. I talk about my own practice (yes, AI images sometimes; yes, Grammarly; no, not the writing or the thinking), about teaching at university in the middle of all this, and about why, as an anthropologist, I think this debate is really a debate about authorship and authenticity wearing a technological costume.The question, in the end, was never if AI. It was always, only, how.CONNECT & SUPPORT

Why organizations investing in artificial intelligence without investing in human readiness are setting themselves up for failed adoption and wasted capital In this episode: Dr. Heather Morton, Tom Bradshaw, LindaAnn Rodgers,  Nic Kruegar, Dr. Jagadesh Chander, Cynthia Cole     I/O Career Accelerator Course: https://www.seboc.com/job Visit us https://www.seboc.com/ Follow us on LinkedIn: https://bit.ly/sebocLI Join an open-mic event: https://www.seboc.com/events   References: 2026 Global Human Capital Trends. (2026). Retrieved from https://www.deloitte.com/us/en/insights/topics/talent/human-capital-trends.html?id=us%3A2ps%3A3gl%3Achc26%3Aawa%3Agreendot%3Anonem%3AK0219492%3A041326%3Akwd-461493411780%3A196756941353%3A804670332474%3A%3A&gclsrc=aw.ds&gad_source=1&gad_campaignid=23331251908&gbraid=0AAAAADenGPBvHbAOgkq78Jysdo7i1C0lG   Crumley, B. (2026, February, 25). Businesses bet big on AI while cutting workers—economists say that's produced ‘zero growth'. Inc.  https://www.inc.com/bruce-crumley/businesses-bet-big-on-ai-while-cutting-workers-economists-say-thats-produced-zero-growth/91307695   Daly, S. J., Wiewiora, A., & Hearn, G. (2025). Shifting attitudes and trust in AI: Influences on organizational AI adoption. Technological Forecasting and Social Change, 215, 124108. Lee, C. P., Lee, M. K., & Mutlu, B. (2026). Making the Invisible Visible: Understanding the Mismatch Between Organizational Goals and Worker Experiences in AI Adoption. arXiv preprint arXiv:2605.03078.   McKinsey & Company. (2025, January 2025). AI in the workplace. A report for 2025.  https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work   Ratanjee, V. (2026, February 14). The deficit reflex: Why change management keeps solving the wrong problem. Forbes.  https://www.forbes.com/sites/vibhasratanjee/2026/02/12/the-deficit-reflex-why-change-management-keeps-solving-the-wrong-problem/

This episode was recorded in March 2026. Gudrun speaks again with Nadja Klein and Moussa Kassem Sbeyti who work at the Scientific Computing Center (SCC) at KIT in Karlsruhe. As a new person in our conversation we welcome Nicolas Bianco. The research of the scientists in Nadja's MBD Lab is at the intersection of statistics and machine learning. It spans theoretical analysis, method development and real-world applications. Last time we focussed on Baysian statistics. With the help of Nicolas we want to examplify how interdisciplinary work is done and how his journey led him into this field of research. Since in this episode we very much focussed on Nicolas decision process and steps in his carrier we plan to have an episode on the topics later in the year. Literature and further information Webpage of the group Webpage of Moussa Kassem Sbeyti Webpage of Nicolas Bianco M. K. Sbeyti and N. Klein: Depth as prior knowledge for object detection, arXiv:2602.05730, 2026 Podcasts Ep 42 of Vanishing Gradient, Jan 2025. O. Beige, G. Thäter: Risikoentscheidungsprozesse, Gespräch im Modellansatz Podcast, Folge 193, Fakultät für Mathematik, Karlsruher Institut für Technologie (KIT), 2019. N. Klein, M. Sbeyti, G. Thaeter: Bayesian Learning, Gespräch im Modellansatz Podcast, Folge 253, Fakultät für Mathematik, Karlsruher Institut für Technologie (KIT), 2025.

Plus a quietly extraordinary week of Portland startup news on top of that. Hydrolix is named Portland's fastest-growing company at 2,721% two-year revenue growth — with nearly flat headcount. Tellagence drops LLM variability from 22–28% down to 1–2.5% with a freshly-published arXiv paper. Prophetic lands M/I Homes, its second national homebuilder after DR Horton. The Portland Metro Region Innovation Hub is searching for yet another executive director. Enduring Planet closes a $12 million climate fund. And the usual secrets.CHAPTERS:00:00 Oregon startup news01:30 ⁨@RecklessBen⁩ Bricks & Minifigs04:50 Portland Metro Region I˜nnovation Hub10:45 Prophetic13:30 Hydrolix16:15 SecretsLINKS:Portland Business Journal names Hydrolix the fastest growing startup in the region — https://siliconflorist.com/2026/06/05/portland-business-journal-names-hydrolix-the-fastest-growing-startup-in-the-region/Hydrolix on PBJ — https://www.bizjournals.com/portland/news/2026/06/05/hydrolix-portland-fastest-growing-companies.htmlTellagence drops LLM variability from 22–28% to 1–2.5% — https://siliconflorist.com/2026/06/04/tellagence-drops-llm-variability-from-22-28-to-1-2-5/Tellagence — https://www.tellagence.ai/Portland startup Prophetic lands another national homebuilder: M/I Homes — https://siliconflorist.com/2026/06/04/portland-startup-prophetic-lands-another-national-homebuilder-m-i-homes/Prophetic — https://www.propheticsoftware.ai/Portland Metro Region Innovation Hub loses another executive director — https://siliconflorist.com/2026/06/02/portland-metro-region-innovation-hub-loses-another-executive-director/Portland Metro Region Innovation Hub — http://www.portlandmetrohub.org/Enduring Planet closes $12M climate focused fund — https://siliconflorist.com/2026/06/03/enduring-planet-closes-12m-climate-focused-fund/The Oregon angle on the Reckless Ben v Bricks & Minifigs drama — https://siliconflorist.com/2026/06/04/the-oregon-angle-on-the-reckless-ben-v-bricks-minifigs-drama-thats-overtaking-geeky-social-media/FIND RICK TUROCZY ON THE INTERNET AT…- https://patreon.com/turoczy- https://linkedin.com/in/turoczy- Portland Oregon startup news on Apple Podcasts https://podcasts.apple.com/us/podcast/portland-oregon-startup-news-silicon-florist/id1711294699- Portland Oregon startup news Spotify https://open.spotify.com/show/2cmLDH8wrPdNMS2qtTnhcy?si=H627wrGOTvStxxKWRlRGLQ- Startup Stories on Spotify https://open.spotify.com/show/1Tk7bbzaNYowGouI9ucKC3- Startup Stories on Apple Podcasts https://podcasts.apple.com/us/podcast/startup-stories-with-silicon-florist/id1849468494- The Long Con on Apple Podcasts https://podcasts.apple.com/us/podcast/the-long-con/id1810923457- The Long Con on Spotify https://open.spotify.com/show/48oglyT5JNKxVH5lnWTYKA- https://bsky.app/profile/turoczy.bsky.social- https://siliconflorist.substack.com/- https://pdxslack.comABOUT SILICON FLORIST ----------For nearly two decades, Rick Turoczy has published Silicon Florist, a blog, newsletter, and podcast that covers entrepreneurs, founders, startups, entrepreneurship, tech, news, and events in the Portland, Oregon, startup community. Whether you're an aspiring entrepreneur, a startup or tech enthusiast, or simply intrigued by Portland's startup culture, Silicon Florist is your go-to source for the latest news, events, jobs, and opportunities in Portland Oregon's flourishing tech and startup scene. Join us in exploring the innovative world of startups in Portland, where creativity and collaboration meet.https://siliconflorist.com#pdx #portland #oregon #startup #entrepreneur

Send us Fan MailCan AI copilots really keep up with pathologists when the cases are new, the workflow is messy, and the benchmark is actually protected from leakage?In this episode of DigiPath Digest #48, I focus on one paper: DALPHIN: Benchmarking Digital Pathology AI Copilots Against Pathologists on an Open Multicentric Dataset. I chose this paper because I think the field needs more of this kind of work. Less hype. More evaluation. Less “look what AI can do.” More “how do we test it in a way that actually means something?” In this session, I look at what makes DALPHIN important for pathologists, lab leaders, and digital pathology trailblazers trying to make sense of pathology AI right now. The paper benchmarks three models against human pathologists: two general-purpose models, Gemini 2.5 Pro and GPT-5, and one pathology-specific model, PathChat+. The dataset includes 1,236 images from 300 cases, covering 130 diagnoses, 14 pathology subspecialties, and cases from six countries. Human performance is benchmarked with 31 pathologists from 10 countries. What I like about this paper is that it does not stop at top-line performance. It deals with the benchmarking problem itself. The authors built a sequestered, indirectly accessible ground truth so the evaluation data could not simply be scraped into model training. That matters because without that protection, benchmarking can become an illusion of genius rather than a real test of generalization. The results are interesting and more nuanced than a simple win-or-lose story. PathChat+ reached expert-level performance in four of six tasks, Gemini in two of six, and GPT in one of six. That tells us something important already: pathology-specific training matters. But it also does not mean pathology is solved. In organ recognition, expert pathologists still outperformed all the models. In rare cancers, none of the models reached expert-level performance. And in ambiguous cases, the models still struggled with something human pathologists do all the time: expressing uncertainty. I also spend time on one of the most practical parts of the paper: model behavior. Gemini tended to overcall. GPT tended to undercall. PathChat was more balanced. That matters in practice. A pathologist using a copilot needs to know the tool's calibration bias before they can safely interpret what it is telling them. I also talk about anchoring bias in conversational interfaces, where early hallucinations can propagate through later answers if memory is not reset between questions. That is not just a technical curiosity. That is a workflow and safety issue. Why should you listen? Because this episode is really about a bigger question: What kind of evidence should pathologists demand before AI copilots enter real workflows? If you want to understand validation, data leakage, rare-case performance, uncertainty, and why these tools should still be treated as co-pilots rather than autopilots, this is a useful paper to know. Episode Highlights01:20 – Why I chose the DALPHIN preprint and why benchmarking matters right now. 05:38 – What is in the DALPHIN dataset: 300 cases, 130 diagnoses, 14 subspecialties, 6 countries. 07:57 – Top-line performance: PathChat+ reaches expert-level performance in 4 of 6 tasks. 09:41 – The benchmarking trap of data leakage and why DALPHIN's sequestered ground truth matters. 12:19 – Why real pathology diagnosis is not text-only and why macro + micro context matters. 15:26 – Tissue recognition, neoplasm detection, ambiguity, and conversational memory: how the testing was structured. 21:29 – The diagnostic personalities of the models: overcalling, undercalling, and balanced behavior. 24:36 – Rare cancers: where AI copilots still fall short of expert human performance. 28:00 – Why binary outputs are not enough when pathology often lives in uncertainty. 31:37 – Anchoring bias and conversational memory: how early hallucinations can keep propagating. 37:11 – Why these tools should be treated as co-pilots, not autopilots. 40:29 – Resources for beginners: Digital Pathology 101 and continued AI literacy. Resources mentionedDALPHIN preprint: arXiv:2605.03544v1 DALPHIN evaluation platform: dalphin.grand-challenge.org PathChat+ pathology-specific AI model discussed in the benchmark. Digital Pathology 101 free eBook by Dr. Aleksandra Zuraw. Educational streams on tissue recognition and computer vision literacy mentioned in the session.Support the showGet the "Digital Pathology 101" FREE E-book and join us!

In this video I talk to YY Ahn about using Claude Code for Research. Normally in Too Lazy to Read the Paper, somebody explains a paper to me, because reading papers is hard and talking to people is fun and natural.   And if someone's too lazy to read the paper, they're definitely too lazy to read the manual. So today there's a new format. There's no paper. Instead there's a thing we're all talking about in the shadows – agentic coding, Claude Code. The whole business of handing an AI your research and letting it write and run the code – and my sense is that getting started on this is a big hurdle to a lot of folks. So to help them (and myself, I'm sitting down with all-round genious, renessaince man, deep thinker, conneseur of all things coding and living in the terminal, and honest, decent person, YY Ahn.   YY is* Quantitative Foundation Distinguished Professor* School of Data Science* University of VirginiaWe cover the following topics* Intro to YY* Brief intro to Claude Code … what even is it?* Working in the terminal* Claude as reviewer zero* Submitting to arXiv, checking references* Some science use-cases * The concrete setup … remote control, etc. What about context windows, should Claude be re-started once in a while? Do you have tasks that run on a schedule.* What about Gemini or ChatGPT as an alternative. Having interchangble coding engines.* Should one have a local model that the AI can direct.* Cool tips & tricks for making life easier* Discussion about what this does to science, where can humans still make a difference*Discussion about de-skilling

Evénements GESDA/OQI à Genève le 6 maiRencontre organisée par l'Open Quantum Institute sur les questions de gouvernance des technologies quantiques. https://www.linkedin.com/posts/gesda-global_quantumdiplomacy-sciencediplomacy-quantumgovernance-activity-7458144545340440576-nVxe Le quantique au service de l'IA et de la robotique : quels horizons ?Je participais le 12 mai au webinar Le quantique au service de l'IA et de la robotique : quels horizons ? organisé par Meanwhile France, en compagnie d'Amélie Cordier (Graine d'IA), Sacha Stojanovic (Meanwhile) et animé par Mickael Aelbrecht (Micke Science). https://www.youtube.com/watch?v=Su7BhM8iY2o Rencontre quantique à l'Ambassade d'AllemagneLe même jour avait lieu une rencontre à l'Ambassade d'Allemagne avec les écosystèmes quantiques français et allemands, regroupant une centaine de personnes. Quandela en a profité pour confirmer ses différents partenariats en Allemagne, avec Attocube et aussi Munich où ils ont une cleanroom de fabrication de semiconducteurs.https://www.quandela.com/fr/newsroom-posts/franco-german-quantum-technology-cooperation/ QEI workshop à BarceloneIl avait lieu du 18 au 22 mai et rassemblait 112 participants et 40 intervenants, couvrant les mondes académiques et industriels. A noter les interventions de Gerard Milburn (NQCC, UK), Natalia Ares (Oxford), Chris Langer (Quantinuum), Christiane Koch (Berlin), Fernando Brendao (Amazon), et Michael Vasmer (Inria). Le résumé est sur LinkedIn et les vidéos seront bientôt disponibles.https://www.linkedin.com/posts/ezratty_the-third-quantum-energy-initiative-workshop-activity-7463937587116888064-DVNi?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAAVgdUB3BosVI5xO_hehYGjYCuwBq1TXaA Llivre blanc Quantum computing in the net-zero transition: energy production, management, and efficiency publié par l'European Energy Research Alliance (EERA), Mai 2026 (30 pages).https://www.eera-set.eu/component/attachments/attachments.html?task=view&id=2178 Q-Expo à BilbaoCet événement organisé par QuIC avait lieu du 18 au 20 mai. https://qexpo.org/ Annonce du renouvèlement de la stratégie nationale quantique le 22 mai Le 22 mai 2026, le Président de la République annonçait le renouvèlement du plan quantique de 2021 avec un financement d'un milliard d'euros. C'était accompagné d'un investissement de 550 millions d'Euros pour la filière des semiconducteurs. L'annonce avait lieu au TGCC du CEA à Bruyères-le-Chatel.   Au passage, Nvidia a investi dans la startup Alice &Bob pour plusieurs dizaines de millions d'euros, mais le montant n'est pas public, l'implantation par Quandela de sa future usine de semi-conducteur à Munich. ColibriTD était aussi mis en avant dans les panels de la journée et annonçait une levée de fonds de 4 M€ menée par le fonds allemand Earlybird et la participation à un consortium aux côtés de MDU (Suède), l'ENAC (France) et LOBA (Portugal) avec un budget global d'environ 1 M€ (dont 255 k€ pour ColibriTD) pour développer des solutions d'IA quantique pour optimiser la gestion du trafic aérien.  Inauguration de la ligne pilote Champ'ionle 26 mai lors de cette inauguration à Villach en Autriche en compagnie d'intervenants Allemands et Autrichiens comme Thomas Monz d'AQT. https://www.linkedin.com/posts/quantumtechnology-quantumcomputing-iontrap-ugcPost-7465341447311876096-hxy9/?utm_source=share&utm_medium=member_ios&rcm=ACoAAAAVgdUB3BosVI5xO_hehYGjYCuwBq1TXaA A venir :Journée en l'honneur de Philippe Grangier à l'IOGS le 4 juin. Voici le lien de retransmission en live.https://youtube.com/live/7aQScUtzvdEFrance Quantum le 16 juin. Billeets gratuits offert par OVHcloud https://www.francequantum.fr/content/full-day-standard-ticket-2026?discount=OVHCLOUDFQ&productid=a1a41d57-08d5-f011-8195-0022487f0371&qty=1Vivatech Panorama de toutes les voies technologiques de l'ordinateur quantique les 25 et 26 juin à Grenoble organisé par la Maison du Quantique Grenoble-Alpes. https://quantalps.univ-grenoble-alpes.fr/quantum-technologies-deep-dive-1713265.kjspUne conférence sur la correction d'erreurs fin août à Amsterdam. France Quandela Ils publiaient en compagnie de la société allemande de « contract research » Walrus Computingun préprint d'estimation de ressources pour exécuter un algorithme de simulation d'un modèle de Heisenberg avec leur modèle FTQC SPOQC. Two Layers, No Swaps: Biplanar SPOQC Architecture Improves Runtime of Fermi-Hubbard Simulation by Boris Bourdoncle, Peter-Jan Derks, Théo Dessertaine, and Johannes Frank, arXiv, May 2026 (44 pages). Et, évolutions dans le management. Cyril Dujardin devient Directeur des Opérations, Michel Zecri vice-Président Industrialisation et Michel Paulin (ex-DG d'OVHcloud) est nommé chairman. Cyril Dujardin « supervisera les opérations mondiales de Quandela et accompagnera l'exécution de la stratégie de l'entreprise dans les domaines du produit, de la R&D, du développement commercial, des partenariats et des déploiements clien...

Fault Tolerance for Quantum Inputs and Outputs with Matthias ChristandlWhy This Episode MattersMost discussions of fault tolerance quietly assume a classical-in, classical-out picture: you feed in bits, the noisy quantum machine does its work, and a stable classical answer comes out the other side. Christandl — a mathematically trained quantum information theorist who also leads a Novo Nordisk Foundation–funded life sciences center — argues that this framing is too narrow for the era we are actually entering, where multi-core processors, networked QPUs, and quantum communication links all need to exchange quantum information between noisy machines.If you care about how quantum networks, distributed quantum computers, and quantum simulation workflows for chemistry and biology actually get built, this episode lays out a foundational way of thinking about the problem and connects it directly to current hardware and algorithm co-design.SponsorThis episode is brought to you by Outshift, Cisco's incubation engine. The need for computational power is rapidly increasing in every sector. From drug discovery to material innovation to complex financial modeling, classical systems are reaching their absolute limits. It's time for a paradigm shift. The answer is a scalable quantum network, built on open standards and vendor-agnostic architecture. By uniting distributed quantum devices, you unlock limitless computational power. Learn more about the Cisco Universal Quantum Switch at Outshift.com.Go deeper with the blog post.What We Get IntoWhy the fault tolerance theorem as usually stated leaves out the case that matters most for networking: quantum inputs and quantum outputs.How Christandl's group shows you can still prepare arbitrarily complex quantum states on a noisy machine, paying only one final layer of physical noise rather than collapsing the whole computation.What this means for restoring meaning to quantum channel capacity results in the presence of noisy encoders and decoders.Why distributed quantum computing — multi-core QPUs talking to each other in quantum, not classical, information — is the natural setting for this work.How recent quantum LDPC code work fits in, and why the team is now focused on making encoders and decoders more space-efficient.Christandl's debate with Gil Kalai: which skeptical assumptions are worth taking seriously, and which he thinks the fault tolerance machinery is robust against.The Quantum for Life workflow: zooming in on the quantum-relevant region of a protein–ligand interaction, running a small quantum simulation, and feeding the result into a classical machine-learning pipeline that needs many such small computations.Why "co-design" has replaced "bridging the gap" as the right metaphor for where quantum hardware and quantum software meet.How quantum sensing — for example, magnetic-field sensing with atomic clouds — could one day deliver genuine quantum inputs into a fault-tolerant quantum computer.Resources & LinksGuest LinksMatthias Christandl — University of Copenhagen Research Portal — Official institutional profile with publications and affiliations.Quantum for Life Center — University of Copenhagen — The Novo Nordisk Foundation–funded center Christandl leads, focused on quantum algorithms for the life sciences.UCPH Quantum Hub launch — The cross-faculty quantum community Christandl helped found at the University of Copenhagen.Christandl appointed 2024 Turing Chair — CWI/QuSoft — Background on his honorary visiting chair at QuSoft and CWI in Amsterdam.Papers & ArticlesFault-Tolerant Coding for Quantum Communication (arXiv:2009.07161) — The foundational paper (IEEE TIT 2024, with Müller-Hermes) that motivates the episode: channel coding when the encoder and decoder circuits themselves are noisy.Fundamental Limit on the Power of Entanglement Assistance in Quantum Communication (arXiv:2408.17290) — Christandl and collaborators settle a 2002 conjecture of Bennett et al. on entanglement-assisted capacity (PRL 2025).Asymptotic tensor rank is characterized by polynomials (arXiv:2411.15789) — STOC 2025 result connecting tensor theory to the matrix multiplication exponent.How to Use Quantum Computers for Biomolecular Free Energies (2026)More Quantum Chemistry with Fewer Qubits — Physical Review Research (2024) — The Quantum for Life paper underlying the protein–ligand workflow discussed in the episode.A Cornerstone of Entanglement Theory Restored — Nature Physics (2025) — Christandl's News & Views on the re-proof of the generalized quantum Stein's lemma.Quantum Duel: Matthias Christandl x Gil Kalai Key Quotes & InsightsOn reframing fault tolerance: Christandl argues that the fault tolerance theorem, as usually stated, assumes classical inputs and outputs — but the most important near-term use cases, from networked QPUs to multi-core processors, need quantum inputs and quantum outputs.On the unavoidable final layer of noise: "There will always be a final layer of noise being applied" when a noisy machine prepares a quantum state — and that single layer, not the whole computation, is the real price you pay.On the new metaphor: "A few years back, I would have told you the really important thing is bridging the gap between the hardware and the software. Now it's not anymore about bridging the gap. It's about working together."On Kalai's skepticism: Christandl finds the debate clarifying rather than threatening — the fault tolerance techniques look robust to the noise-model perturbations skeptics raise, and the engineering question is which code, not whether codes work at all.On what quantum advantage in life sciences might actually look like: Not one heroic simulation, but many small, exact quantum computations feeding training data into a much larger classical machine-learning workflow that predicts protein–ligand interactions.Related Episodes

OpenAI debuted personal finance tools via Plaid for Pro users. AI startups generate ~$80B in annualized revenue, with Anthropic and OpenAI capturing 89%. ArXiv cracks down on AI slop, Apple's Siri relaunch may still be a beta, and SF vibes are frenetic. OpenAI debuts personal finance tools for US ChatGPT Pro users, partnering with Plaid to give access to 12K+ financial institutions to analyze spending and more (TechCrunch) Analysis: 34 leading AI startups are generating ~$80B in annualized revenue, up 112% from six months ago, with Anthropic and OpenAI capturing 89% of the revenue (The Information) ArXiv, the repository of preprint academic research, says it will ban authors for a year if their papers have "incontrovertible evidence" of AI-generated work (404 Media) Sources: Apple's revamped Siri may launch in beta, and will have an option to auto-delete chats; Apple plans to add Suggested Genmoji to iOS 27 and iPadOS 27 (Bloomberg) A college senior at Stanford describes how AI has changed classes: cheating using AI "has become omnipresent" with students "fudging just about everything" (NYT) SF vibes are frenetic over the huge divide in outcomes and career uncertainty for software engineers; over 5 years ~10K people in AI attained retirement wealth (X / @deedydas) Learn more about your ad choices. Visit megaphone.fm/adchoices

Welcome to Nerd Alert, a series of special episodes bridging the gap between marketing academia and practitioners. We're breaking down highly involved, complex research into plain language and takeaways any marketer can use. In this episode, Elena and Rob explore how AI shopping agents make purchase decisions and what the results mean for brands that aren't optimized to be found, chosen, or endorsed by an AI. Topics covered: [01:45] "What Is Your AI Agent Buying? Evaluation, Biases, Model Dependence, and Emerging Implications for Agentic E-Commerce"[05:30] AI agents pile onto winner products and ignore everyone else[07:00] Which AI model does the shopping changes everything[08:00] How models respond differently to product position on the page[09:00] Why "Overall Pick" badges dramatically boost selection rates[09:45] How rewriting a product title drove market share from 0% to 41% To learn more, visit marketingarchitects.com/podcast Resources: Allouah, A., Besbes, O., Figueroa, J. D., Kanoria, Y., & Kumar, A. (2025). What is your AI agent buying? Evaluation, biases, model dependence, & emerging implications for agentic E-commerce. https://doi.org/10.48550/arXiv.2508.02630  Get more research-backed marketing strategies by subscribing to The Marketing Architects on Apple Podcasts, Spotify, or wherever you listen to podcasts.  

Hardware-Faithful Digital Twins for Quantum Computing with Izhar MedalsyIzhar Medalsy is not a career qubit theorist. His path runs from a physical chemistry PhD and an ETH Zurich postdoc in atomic force microscopy and ternary nanoscale logic, through productizing scientific instruments at Bruker, through building one of the fastest resin 3D printers on the market, into co-founding Quantum Elements in 2023 with Daniel Lidar (USC) and Amir Yacoby (Harvard). That arc — nanoscale measurement scientist turned deep-tech operator — shapes how he thinks about the simulation gap in quantum computing.The conversation lands at a specific moment. In April 2026, Quantum Elements published a joint result with AWS, USC, and Harvard simulating a distance-7 rotated surface code with 97 physical qubits using full quantum master equations on AWS HPC7a, and announced a deeper collaboration with Rigetti Computing on next-generation superconducting processors. If you care about how error correction strategies, decoders, and pulse-level controls actually get developed before they ever touch hardware, this episode is for you.EPISODE SPONSORThis episode is brought to you by Outshift, Cisco's incubation engine. The need for computational power is rapidly increasing in every sector. From drug discovery to material innovation to complex financial modeling, classical systems are reaching their absolute limits. It's time for a paradigm shift. The answer is a scalable quantum network, built on open standards and vendor-agnostic architecture. By uniting distributed quantum devices, you unlock limitless computational power.Learn more about the Cisco Universal Quantum Switch at Outshift.comGo deeper with the blog post The switch that quantum networking has been waiting for====================================================================================================What We Get IntoWhy generic noise models fall short and what "hardware-faithful" actually means when two nominally identical QPUs have different noise fingerprintsHow Quantum Elements scaled open-system master-equation simulation from a brute-force ceiling around 16 qubits to 97 qubits using stochastic compression on top of Quantum Monte CarloThe compute reality of the distance-7 surface code run on AWS HPC7a — only 96 vCPUs and a few hundred gigabytes of memory, not the thousands of vCPUs they initially fearedWhy decoders are the invisible bottleneck in fault tolerance, and where AI-trained decoders fed by digital twin data could plausibly run inside the real-time quantum-classical loopExtending error suppression from physical qubits up to logical qubits — the IBM Eagle work where digital-twin-guided strategies reportedly took entangled logical qubit fidelity from 43% to 95%How the same digital twin approach extends to neutral atoms (live today) and ion traps (on the roadmap)What Rigetti gets out of the partnership, what it means to have Chad Rigetti on the board, and how Constellation fits alongside real hardware timeIzhar's "wooden models in the air tunnel" critique of how the quantum industry currently iterates — and what a parallel virtual development track buys youResources & LinksGuest & CompanyIzhar Medalsy — Quantum Elements team page — Background and role at Quantum Elements.Izhar Medalsy on LinkedIn — Full career arc from ETH biophysics through 3D printing to quantum.Quantum Elements — Constellation platform, where listeners can build their own virtual QPU and run circuits, error suppression, and QEC experiments.Papers & ArticlesAWS Quantum Computing Blog: Decoding realistic QEC syndrome with Quantum Elements digital twins — Primary technical reference for the 97-qubit distance-7 result discussed in the episode.The Next Platform: How HPC and AI Digital Twins Accelerate Quantum Error Correction (Apr 17, 2026) — Independent reporting on the AWS/USC/Harvard simulation.The Quantum Insider: Quantum Elements & Rigetti collaboration (Apr 21, 2026) — Details on the partnership Izhar describes.Guest post: Quantum Digital Twins — The Missing Acceleration Layer — Izhar's own framing of the thesis.The Next Platform: Startup Profile of Quantum Elements (Jan 2026) — Background on the company.arXiv 2603.14607 — Calibration-Based Digital Twins for IBM Quantum Hardware — Useful independent context on the limits and promise of calibration-based twins.Key Quotes & Insights"Sometimes when I look at the quantum industry, there are instances where you think, well, it's almost like building the next fighter jet with wooden models in the air tunnel." — Izhar's framing for why the field needs a real simulation layer.On hardware awareness: each modality, each QPU, sometimes each calibration cycle has its own pulses, its own noise processes, and its own failure modes. You cannot build the control stack without modeling where you are starting from and where you are trying to get to.Insight: The brute-force ceiling for open-system master-equation simulation is roughly 16 qubits. Stochastic compression layered on Quantum Monte Carlo is what let Quantum Elements reach distance-7 surface code at 97 qubits — exploiting sparsity rather than enumerating the full state space.On logical qubits: "We cannot assume that logical qubits will be noise-free." Error suppression strategies developed at the physical level need to be re-derived at the logical level, and digital twins are how you train and test those strategies before hardware.Insight: The most interesting downstream story may not be simulation itself but AI decoders trained on digital-twin-generated data — small enough to run at the edge, fast enough to live inside the real-time quantum-classical loop.Related EpisodesEpisode 52 — Quantum noise with Daniel Lidar — Quantum Elements' co-founder and CSO on the noise suppression and error correction foundat...

How do you go from a quantum computing research paper to a circuit you can run on a real machine? Can AI help bridge the gap between complex mathematical theory and functional quantum code? In this episode, Vincent van Wingerden, Director of Strategic Partnerships at Classiq, joins host Konstantinos Karagiannis to discuss the dawn of quantum vibe coding. While traditional quantum programming often requires tedious manual manipulation of gates and hardware-specific constraints, Classiq's platform introduces a high-level abstraction that allows users to focus on intent rather than implementation. Vincent explains how their AI-driven assistant, powered by Claude, can ingest an academic paper from arXiv and generate a working implementation in minutes, a task that once took experts weeks of effort.  Beyond the vibe, this episode dives into the practical mechanics of how Classiq's compiler optimizes code across diverse hardware modalities, from IonQ to IBM. You'll hear about real-world collaborations with Comcast and AMD, as well as a fascinating example of how an AI agent suggested improvements to an image-edge-detection algorithm that outperformed its original human design. Whether you are a seasoned computer scientist or a curious beginner looking to bypass the gate-level barrier, this conversation reveals how the synergy between high-level languages and AI is dismantling the quantum talent gap and scaling the industry toward the future.  For more information on Classiq, visit www.classiq.io/.   An example of quantum code aided by AI https://github.com/Classiq/classiq-library/tree/main/applications/image_processing/quantum_hadamard_edge_detection.Visit Protiviti at www.protiviti.com/US-en/technology-consulting/quantum-computing-services to learn more about how Protiviti is helping organizations get post-quantum ready.  Follow host Konstantinos Karagiannis on all socials: @KonstantHacker             Questions and comments are welcome!  Theme song by David Schwartz, copyright 2021.  The views expressed by the participants of this program are their own and do not represent the views of, nor are they endorsed by, Protiviti Inc., The Post-Quantum World, or their respective officers, directors, employees, agents, representatives, shareholders, or subsidiaries.  None of the content should be considered investment advice, as an offer or solicitation of an offer to buy or sell, or as an endorsement of any company, security, fund, or other securities or non-securities offering. Thanks for listening to this podcast. Protiviti Inc. is an equal opportunity employer, including minorities, females, people with disabilities, and veterans.  

“The truth is paywalled, and the lies are free.” — Current Affairs editor, quoted by Brewster Kahle The internet, we were promised, would remember everything. Rather than memory, however, it is now most distinguished by its digital forgetfulness. That's the warning in Vanishing Culture, a new series of essays published by the San Francisco-based Internet Archive. In its concluding essay by Brewster Kahle — founder of the Internet Archive, member of the Internet Hall of Fame, and the closest thing the web has to an official librarian — he makes the case for preserving the online library system. “Our evolving digital age can be our next Carnegie moment or it can be a Library of Alexandria moment. It is up to us.” Today's internet library system, Kahle argues, is worse than the analogue one he grew up with. It's faster, he acknowledges, but shallower. The 1976 Copyright Act means that rather than buying digital books, libraries can only rent access in surveillance environments controlled by a handful of corporations. Sixty percent of news organisations now have paywalls. Academic publishing is controlled by three conglomerates. So an entire generation is growing up without access to the published works of the twentieth century. “The truth is paywalled, and the lies are free,” as the editor of Current Affairs put it. That is today's internet. No laughter. Only forgetting.Five Takeaways •       Carnegie Moment or Alexandria Moment: The Internet Archive's pamphlet Vanishing Culture opens with a choice. Andrew Carnegie invested in public libraries during the early twentieth century: every town in America got one, and by the time the US was thrust onto the world stage after World War II, an educated public was ready. The Library of Alexandria burned. Kahle's argument: we are at the same fork in the road. The digital transition can be a Carnegie moment — everyone with access to all human knowledge — or it can be an Alexandria moment. Sixty percent of news organisations now have paywalls. Academic publishing is controlled by three conglomerates. The library system we have is worse, not better, than the one Kahle grew up with. •       The 1976 Copyright Act as Original Sin: Copyright used to be opt-in: you had to put a ‘c' on your work and register it. The 1976 Act made it opt-out: everything is copyrighted by default, forever, with terms that keep being extended. The consequences: Wikipedia had to be written from scratch because the encyclopedias already written couldn't be shared openly. Academic papers are walled inside publisher systems, which is why arXiv exists. Libraries can no longer buy digital books — only rent access in surveillance environments. The bargain between publishers, libraries, authors, and the public that functioned for centuries has been dissolved by lobbyists writing copyright law. •       The Truth Is Paywalled and the Lies Are Free: Kahle's most quotable line belongs to someone else — the editor of Current Affairs. But Kahle endorses it fully. An entire generation is now growing up without access to the published works of the twentieth century. People are genuinely confused about whether the Holocaust happened — not because the information doesn't exist, but because it's behind a paywall. What is free on the internet is what serves the interests of the platforms: viral, emotional, algorithmically optimised, frequently false. The deep, sourced, accurate record costs money to access. That inversion is not an accident. It is the business model. •       Turnkey Tyranny: Kahle quotes Edward Snowden's phrase for what surveillance capitalism has built: turnkey tyranny. All it needs is someone motivated to think tyrannically, and all the laws, policies, and technologies are already in place. The internet was built on a protocol: play by the rules and you're in. That openness is gone. What replaced it is a small number of platforms with enormous centralised control of distribution, purchasing the upstream sources — Comcast buying movie studios, Amazon buying MGM. Whoever controls distribution, Lawrence Lessig's maxim holds, will eventually control everything upstream from it. •       AI Mass Larceny? The Real Loser Is People: Asked the binary question — is AI mass larceny, yes or no? — Kahle refuses it. His answer: the fight between publishers and AI companies is Coke versus Pepsi. The real dynamic is large corporations — whether you call them AI companies or publishing conglomerates — taking from people's goodwill, their creative output, their authorship, and landing the value in very few hands. What Kahle wants is public AI: ClimateGPT, reading the Sri Lankan 1953 fish reports and seeing the patterns in them. AI that serves the public good, not the shareholders of one, two, or three gigantic players. The answer isn't either Coke or Pepsi. It's water. About the Guest Brewster Kahle is the founder and Digital Librarian of the Internet Archive, a member of the Internet Hall of Fame, and the author or editor of Vanishing Culture (Internet Archive, 2024). He was previously the founder of WAIS and Alexa Internet. He lives in San Francisco. References: •       Internet Archive — archive.org. •       Vanishing Culture: A Report on Our Disappearing Digital Heritage, ed. Brewster Kahle et al. (Internet Archive, 2024). Available free at archive.org. •       arXiv (arxiv.org) — the open-access preprint server that routes around academic publishing. •       Episode 2877: Keith Teare — Let's Just Say It Out Loud: AI Is Not Dangerous. The counterpoint to Kahle's wariness about AI centralisation. About Keen On America Nobody asks more awkward questions than the Anglo-American writer and filmmaker Andrew Keen. In Keen On America, Andrew brings his pointed Transatlantic wit to making sense of the United States — hosting daily interviews about the history and future of this now venerable Republic. With nearly 2,900 episodes since the show launched on TechCrunch in 2010, Keen On America is the most prolific intellectual interview show in the history of podcasting. WebsiteSubstackYouTubeApple PodcastsSpotify Chapters: (00:30) - The internet's librarian: forgetting vs. surveillance (01:55) - Carnegie moment or Alexandria moment? (03:20) - Andrew Carnegi...

It's once again 'Bring Your Paper to Work' day here at Mindtools Kineo, as Ross G, Dr Anna and Ross D each take turns to share an academic study that they think has key insights for L&D professionals. In this week's episode of The Mindtools L&D Podcast, we discuss the following papers: Rogelberg, S. G., Kreamer, L. M., & Gray, J. (2026). 'Thirty years of meeting science: Lessons learned and the road ahead.' Annual Review of Organizational Psychology and Organizational Behavior, 13, 415–442. Castro, S., Englmaier, F., & Guadalupe, M. (2024). 'Fostering psychological safety in teams: Evidence from an RCT'. Academy of Management Proceedings, 2024(1), 16624. Shen, J. H., & Tamkin, A. (2026). 'How AI impacts skill formation' (arXiv preprint arXiv:2601.20245). In 'What I Learned This Week', Ross D mentioned National Grid: Live. For more from Mindtools Kineo, visit mindtools.com or kineo.com. There, you'll also find details of our Learning Management Systems, Content Hub for leaders and managers, and custom learning design service - including AI skills development! You can also email us at custom@mindtools.com. Like the show? You'll LOVE our newsletter! Subscribe to The L&D Dispatch at lddispatch.com Connect with our speakers     If you'd like to share your thoughts on this episode, connect with us on LinkedIn:  Ross Dickie  Anna Barnett  Ross Garner  

Artificial intelligence is rapidly changing how people search for information, including how families learn about pregnancy and birth. As this technology grows, it raises important questions about accuracy, ethics, and the role of human expertise in healthcare information.   In this episode, Dr. Rebecca Dekker and Dr. Sara Ailshire share a behind-the-scenes look at how Evidence Based Birth® developed its own AI policies. They discuss concerns about misinformation, bias, privacy, and environmental impact, as well as the potential effects on critical thinking and human connection.   Learn why EBB has chosen to avoid the use of AI in our research, and what that means both for our team and for you.   (03:22) Why EBB began developing an AI policy  (10:39) Defining AI, generative AI, LLMs, and hallucinations  (17:03) Ethical concerns: Accuracy, bias, and risks to evidence-based information  (20:14) Environmental impact of AI and data centers  (21:47) Privacy concerns and data security risks  (26:21) Intellectual property, sustainability, and loss of research context  (27:03) Humanitarian concerns and the impact of AI on jobs and communities  (31:24) AI's potential effects on cognition and critical thinking  (37:30) Why EBB takes a cautious, evidence-based approach to AI  (38:20) Research team policy (42:06) Content team policy (45:01) Programs team policy and guidance for applicants  (47:32) Hiring practices and commitment to human review (52:29) Final takeaways: EBB's overall stance on AI    References Read EBB's Statement on Artificial Intelligence: ebbirth.com/artificial-intelligence-statement/ For a Student Who Used AI to Write a Paper, by Joseph Fasano: https://poets.org/poem/student-who-used-ai-write-paper Barrington, F. (2025). "Thirsty for power and water, AI-crunching data centers sprout across the West." & The West Magazine, Stanford University. April 8, 2025. https://andthewest.stanford.edu/2025/thirsty-for-power-and-water-ai-crunching-data-centers-sprout-across-the-west/ Gecker, J. (2025). "Big Tech is paying millions to train teachers on AI, in a push to bring chatbots into classrooms." Associated Press. October 20, 2025. https://apnews.com/article/artificial-intelligence-teacher-union-microsoft-f7554b6550fb90519dd8129acac8e291 Han, Y., Wu, Z., Li, P., et al. (2024). "The Unpaid Toll: Quantifying and Addressing the Public Health Impact of Data Centers." arXiv preprint arXiv: 2412.06288. https://arxiv.org/pdf/2412.06288 Hou, H., Leach, K., & Huang, Y. (2024). "ChatGPT Giving Relationship Advice - How Reliable Is It?" Proceedings of the Eighteenth International AAI Conference on Web and Social Media: 610–623. https://ojs.aaai.org/index.php/ICWSM/article/view/31338 Kosmyna, N., Hauptmann, E., Yuan, Y. T., et al. (2025). "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task." arXiv preprint arXiv:2506.08872. https://arxiv.org/abs/2506.08872 Marrinan, C. (2025). "Data Center Boom Risks Health of Already Vulnerable Communities." Tech Policy Press.com. June 12, 2025. https://www.techpolicy.press/data-center-boom-risks-health-of-already-vulnerable-communities/ NASA. (2024). "Defining Artificial Intelligence." Accessed November 17, 2025. https://www.nasa.gov/what-is-artificial-intelligence/ Notre Dame Learning. (2025). "AI Overview and Definitions." Accessed November 17, 2025. http://learning.nd.edu/resource-library/ai-overview-and-definitions/ Pataranutaporn, P., Karny, S., Archiwaranguprok, C., et al. (2025). "My Boyfriend is AI: A Computational Analysis of Human-AI Companionship in Reddit's AI Community." arXiv preprint arXiv:2509.11391.https://arxiv.org/pdf/2509.11391 Sonka, J. (2025). "The AI data center boom is coming for Kentucky. What will lawmakers do about it?" Kentucky Public Radio. December 9, 2025. https://www.lpm.org/news/2025-12-09/the-ai-data-center-boom-is-coming-for-kentucky-what-will-lawmakers-do-about-it Stryker, C. (n.d.) "What are LLMs?" IBM.com. Accessed November 17, 2025. https://www.ibm.com/think/topics/large-language-models Tabuchi, H. (2025). "Elon Musk's A.I. Company Faces Lawsuit Over Gas-Burning Turbines." New York Times. June 17, 2025. www.nytimes.com/2025/06/17/climate/naacp-musk-xai-supercomputer-lawsuit.html/ United Nations (UN). (2025). "AI has an environmental problem. Here's what the world can do about that." UN Environmental Programme. November 13, 2025. https://www.unep.org/news-and-stories/story/ai-has-environmental-problem-heres-what-world-can-do-about