Podcasts about Lidar

L'autonomie totale des voitures promise par Elon Musk se heurte aujourd'hui à une réalité statistique brutale.Les robotaxis de Tesla, en test au Texas, affichent un taux d'accident quatre fois supérieur à celui d'un conducteur humain moyen. 14 accidents ont été officiellement recensésD'abord, il faut regarder les chiffres de l'expérimentation au-delà du marketing.Depuis huit mois, une flotte de 43 Tesla opérant en mode autonome a parcouru près de 1,3 millions de kilomètres. Et sur cette très longue distance, 14 accidents ont été officiellement recensés.En calculant la moyenne, cela représente une collision tous les 90 000 kilomètres.Pour mettre ce chiffre en perspective, les propres données de Tesla indiquent qu'un conducteur humain moyen ne subit un incident mineur que tous les 368 000 kilomètres.Concrètement, le système de conduite autonome de Tesla est actuellement quatre fois moins sûr que le plus banal des automobilistes texans.La courbe de progression semble s'inverserEnsuite, l'analyse de la nature des accidents révèle des lacunes technologiques inquiétantes pour un déploiement à grande échelle.Les rapports d'accidents font état de crash avec cinq autres véhicules, cinq objets fixes, mais aussi un cycliste et un animal.Plus troublant encore, la courbe de progression semble s'inverser. Alors que l'IA est censée s'améliorer par l'apprentissage continu, plus de 35 % des incidents ont été signalés sur les deux derniers mois de l'étude.Ce constat pose une question fondamentale sur la fiabilité du système vision-only de Tesla, qui refuse d'utiliser les capteurs Lidar, jugés trop chers. Sans une amélioration radicale de ces scores, le passage d'une flotte expérimentale à un service commercial semble s'éloigner.Sincérité ?Enfin, c'est la transparence de Tesla qui est aujourd'hui remise en question.Un incident survenu en juillet dernier, initialement déclaré comme un simple dommage matériel, n'a été requalifié en accident avec hospitalisation que cinq mois plus tard.Ce délai dans la déclaration soulève des interrogations sur le reporting de l'entreprise et la sincérité des promesses de son dirigeant.Surtout, la promesse d'une IA conductrice plus sûre que l'humain reste, pour l'instant, une ambition non vérifiée par les faits.Le ZD Tech est sur toutes les plateformes de podcast ! Abonnez-vous !Hébergé par Ausha. Visitez ausha.co/politique-de-confidentialite pour plus d'informations.

Abertura dos trabalhos na Amorosidade

Today I'm joined by David Wyler, CEO of Jeff Wyler Automotive Family. We dig into why Wyler abandoned traditional corporate management in favor of a “coach” model, how a strict 100-mile acquisition rule protects execution, and why culture—not capital—is the only defensible edge left in consolidation. David also unpacks his massive acquisition of the Midwest Auto Group, the lessons of stagnation during COVID, and what it really takes to scale a family business without losing its soul. This episode is brought to you by: 1. YSM Design - YSM Design, your expert in automotive dealer architecture, helps dealer principals and fixed/ variable ops teams improve the bottom line with EV readiness checks, OEM brand image updates/new-store requests, and expansions or renovations—big or small—powered by instant renderings, immersive 360s, and LiDAR scans that reduce surprises and speed decisions; visit @ here or call 404-249-4555 2. Siro - Siro's AI gives dealerships full visibility into every conversation. It records, transcribes, and analyzes in-person conversations. Proactively flagging compliance issues, missed revenue opportunities, and training gaps. Go to @ https://www.siro.ai/cdg to learn more 3. CDG Recruiting - Hire top dealership talent, fast. From sales managers to GMs and C-suite execs, we've placed over 1,000 roles across auto retail. Ready to scale without the hassle? Visit @ https://www.cdgrecruiting.com/ to get started. Check out Car Dealership Guy's stuff: For dealers: CDG Circles ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://cdgcircles.com/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Industry job board ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://jobs.dealershipguy.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Dealership recruiting ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.cdgrecruiting.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Fix your dealership's social media ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.trynomad.co⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Request to be a podcast guest ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.cdgguest.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ For industry vendors: Advertise with Car Dealership Guy ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.cdgpartner.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Industry job board ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://jobs.dealershipguy.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Request to be a podcast guest ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.cdgguest.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Topics: 08:20 Herb Chambers' advice on buying dealerships everywhere. 18:45 Why every single manager got fired and rehired. 24:20 The game film trick that fixed F&I managers. 25:55 How an internal playbook killed outside training programs. 31:40 The brutal truth about being a second-generation dealer. 39:55 Why practicing golf is actually terrifying. 46:05 What COVID really did to company performance. 46:55 The NFL analogy exposing every dealership's weakness. Car Dealership Guy Socials: X ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠x.com/GuyDealership⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Instagram ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠instagram.com/cardealershipguy/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ TikTok ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠tiktok.com/@guydealership⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ LinkedIn ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠linkedin.com/company/cardealershipguy⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Threads ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠threads.net/@cardealershipguy⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Facebook ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠facebook.com/profile.php?id=100077402857683⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Everything else ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠dealershipguy.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Igal Raichelgauz, Founder & CEO, Autobrains joined Grayson Brulte on The Road to Autonomy podcast to discuss the company's strategic partnership with VinFast and the development of an affordable, scalable robo-car.The operational backbone of Autobrains' strategy is a Thinking AI approach that utilizes an agentic architecture rather than traditional monolithic models. By using a library of specific skills that can be added incrementally, the system scales from basic safety features to full autonomy without requiring massive data retraining or excessive computational power.In the field, Autobrains is rigorously applying its technology to the VinFast VF 8 and VF 9 models, proving the system's robustness in some of the world's most complex driving environments, such as the congested streets of Hanoi, Vietnam. Autobrains utilizes a vision-only approach that mimics human perception to navigate urban traffic, heavy rain, and high-speed highways.Autobrains' Physical AI ecosystem also includes an air to road localization system, which uses compressed satellite imagery signatures to provide 10-centimeter positioning accuracy. Allowing the vehicle to localize itself globally and understand lane boundaries or construction sites without relying on expensive, high-maintenance HD maps.Looking ahead, Igal envisions a future where autonomous driving reaches a mass-market inflection point within the next five years. This evolution aims to fundamentally transform the industry by delivering a fully autonomous robo-car at a $30,000 price point, enabling every vehicle to become a revenue-generating asset that increases safety and gives time back to the consumer.Episode Chapters00:00 How the VinFast Deal Came Together03:16 Skills-Based Agentic AI Architecture 07:16 Six Cameras, 360° Coverage, Low Compute 09:37 Air-to-Road: Satellite Imagery Replaces HD Maps12:40 Robo-car Vision 15:10 The $30K Fully Autonomous Car 20:20 The Thinking Layer24:22 20 Teraflops, Sub-20ms Latency, Edge Computing 27:58 No Lidar: The Vision-Only Thesis 28:59 The Future of Autobrains--------About The Road to AutonomyThe Road to Autonomy is the definitive media brand covering the Autonomy Economy™. Through our podcasts, newsletter, and proprietary market intelligence, we set the narrative for institutional investors, industry executives, and policymakers navigating the convergence of automation, autonomy, and economic growth.Join institutional investors and industry leaders who read This Week in The Autonomy Economy every Sunday. Each edition delivers exclusive insight and commentary on the autonomy economy, helping you stay ahead of what's next. Subscribe today for free: https://www.roadtoautonomy.com/ae/See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Aujourd'hui, nous plongeons dans les coulisses de la conduite autonome avec une percée majeure signée Waymo.La filiale d'Alphabet vient de dévoiler son "Waymo World Model", une intelligence artificielle génératrice de mondes virtuels capable de simuler des situations de conduite avec un réalisme jamais atteint.Ce n'est pas seulement une prouesse technique, c'est le moteur qui va permettre aux véhicules autonomes de franchir un cap critique en matière de sécurité et de passage à l'échelle.Créer des environnements 3D photoréalistes et interactifsConcrètement, ce modèle s'appuie sur Genie 3, l'IA de Google DeepMind, pour créer des environnements 3D photoréalistes et interactifs.Le premier point de rupture, c'est la gestion des cas limites, ce que les ingénieurs appellent le "long-tail".En s'appuyant sur une connaissance du monde apprise via des milliards de vidéos, le simulateur peut inventer des scénarios que la flotte de Waymo n'a jamais croisés dans la réalité, comme une rencontre fortuite avec un éléphant ou une tornade en pleine ville.Là où les simulateurs classiques sont limités par les données collectées sur route, le World Model s'en affranchit donc pour préparer l'IA à l'imprévisible.ContrôlabilitéMais attention, il ne s'agit pas de simples vidéos passives. Le deuxième pilier de cette technologie, c'est la contrôlabilité.Les ingénieurs peuvent modifier une scène par un prompt ou changer la trajectoire du véhicule pour tester des scénarios contrefactuels. Comme par exemple que se serait-il passé si la voiture avait accéléré au lieu de freiner ?L'IA recalcule alors en temps réel non seulement l'image de la caméra, mais aussi les données LiDAR, indispensables pour la perception de la profondeur.C'est cette fusion multi-capteurs qui garantit que ce qui est appris en simulation est directement applicable sur le bitume.Waymo peut transformer n'importe quelle vidéo amateur en une simulation 3DEnfin, la force de ce modèle réside dans sa capacité de conversion.Waymo peut désormais transformer n'importe quelle vidéo amateur ou de dashcam en une simulation 3D.Une rue enneigée filmée par un smartphone devient donc un terrain d'entraînement multi-modal.Associé à une optimisation de l'inférence qui permet de simuler des séquences longues sans explosion des coûts de calcul, Waymo dispose ainsi d'un outil de validation scalable.Au final, la course à l'autonomie ne se gagne plus seulement sur la route, mais dans la capacité à générer et maîtriser des milliards de kilomètres virtuels hyper-réalistes.Le ZD Tech est sur toutes les plateformes de podcast ! Abonnez-vous !Hébergé par Ausha. Visitez ausha.co/politique-de-confidentialite pour plus d'informations.

Não era o tema desta semana, mas devido às circunstâncias (vão perceber) parece-me um tema pertinente... e comum!See omnystudio.com/listener for privacy information.

Soroush Salehian, CEO of Aeva (AEVA), says the latest quarter shows their company is the strongest in the sector for autonomous driving technology. He covers their financials along with their new collaborations, like with Nvidia (NVDA) as a LiDAR supplier. “There's been this race for automation,” he says, explaining how they are expanding from industrial vehicles to passenger vehicles. “We are no longer in the development phase…we are transitioning to scale and deployment, and it's only a couple years away.” Soroush covers also covers the products they create for the defense sector.======== Schwab Network ========Empowering every investor and trader, every market day.Options involve risks and are not suitable for all investors. Before trading, read the Options Disclosure Document. http://bit.ly/2v9tH6DSubscribe to the Market Minute newsletter - https://schwabnetwork.com/subscribeDownload the iOS app - https://apps.apple.com/us/app/schwab-network/id1460719185Download the Amazon Fire Tv App - https://www.amazon.com/TD-Ameritrade-Network/dp/B07KRD76C7Watch on Sling - https://watch.sling.com/1/asset/191928615bd8d47686f94682aefaa007/watchWatch on Vizio - https://www.vizio.com/en/watchfreeplus-exploreWatch on DistroTV - https://www.distro.tv/live/schwab-network/Follow us on X – https://twitter.com/schwabnetworkFollow us on Facebook – https://www.facebook.com/schwabnetworkFollow us on LinkedIn - https://www.linkedin.com/company/schwab-network/About Schwab Network - https://schwabnetwork.com/about

Join us as we sit down with Richard Chelminski, Senior Vice President, Head of SDV Platform, at 42dot, an autonomous driving software and mobility platform development startup from Hyundai Motor Group. We discuss how the company is pushing boundaries with its end-to-end autonomous driving system, Atria AI, which is focused on Level 4 autonomy and designed for seamless integration into consumer vehicles — no LIDAR required! You'll also learn why camera-first and radar technologies are reshaping the industry, how software is now at the heart of vehicle development, and what the future holds for Transportation as a Service (TaaS). If you're curious about the next wave of smart, software-defined vehicles and autonomous driving innovation, this episode is a must-listen.   We'd love to hear from you. Share your comments, questions and ideas for future topics and guests to podcast@sae.org. Don't forget to take a moment to follow SAE Tomorrow Today—a podcast where we discuss emerging technology and trends in mobility with the leaders, innovators and strategists making it all happen—and give us a review on your preferred podcasting platform.   Follow SAE on LinkedIn, Instagram, Facebook, X, and YouTube. Follow host Grayson Brulte on LinkedIn, X, and Instagram.

Thu, 26 Feb 2026 17:15:00 GMT http://relay.fm/conduit/122 http://relay.fm/conduit/122 Kathy Campbell and Jay Miller Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. clean 5583 Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. Guest Starring: Merlin Mann Links and Show Notes: Checked Connections - Merlin ✅ - Working on collecting the old sites and Fives list - Kathy ✅ - Get ready for unicorning cowork Keep sending those MyConduit Connections to us on Discord and through Feedback! New Connections - Merlin - Keep working on the site thing - Kathy - Take things to the post office For Our Super Conductors: Pre-Show: LIDar on iOS. How do you know if you're ladder is against the right wall? Post-Show: Embracing the chaos Credits Music: When You Smile Executive Producers: Relay FM Discord Community Conduit e122 Links Merlin's One Good Things Where Everybody Knows Your Name: Judy Greer (Ted Danson, Conan O'Brien Network) -- "I went in thinking, oh, this looks really good, and I ended up liking it probably twice as much as I expected." Judy Greer -- Cheryl/Carol on Archer, Kitty Sanchez on Arrested Development. "It was neat to hear her talk about how important it was for her to get better at acting." Typora -- WYSIWYG Markdown editor ($15). "A really nice balance of what I'm looking for" -- discovered through the 5ives redesign work with Claude. Judi Dench speech on The Graham Norton Show -- "Made me cry." Kathy's One Good Thing Flavor Flav sponsoring the US women's hockey team -- Vegas celebration for the gold-medal team. Merlin responded by rapping "Bring the Noise" from memory. Merlin's Shows Do By Friday (with Alex Cox) Reconcilable Differences (with John Siracusa) Roderick on the Line (with John Roderick) Productivity / Publishing Inbox Zero -- "I'm the inbox zero guy." Merlin originated the concept; the world turned it into a marketing term. 43folders.com -- "In 2004, there were not a lot of websites about how to deal with your productivity problems as a Mac user." Back to Work (5by5) -- former podcast David Allen / Getting Things Done -- "He claims he's the laziest man in the world, and I've always admired that he says that." Danny O'Brien and the 2005 ETech "Life Hacks" talk -- "Danny and I are both so addled and odd and different... his energy was just incandescent to be around." The conference where Merlin's laptop had Wi-Fi for the first time. Site Meter -- "There's your life before site meter and your life after site meter." The little GIF badge that counted page loads and launched a million blog vanity spirals. 5ives & Typography 5ives -- Merlin's list site (2002), 450 lists, being revived. "I'm pleased with myself. I like that I made four hundred and fifty lists that some people thought were funny in the 2000s." Matthew Butterick -- fonts, Practical Typography. "One of those people where I'm just interested in your deal," like Simon Willison or Edgar Wright. Merlin bought the entire font set during a bout of situational depression and is finally using them for the 5ives redesign. Movies & TV The Hollow Crown (BBC) -- Trailer. "Look at that stacked cast." Ben Whishaw, Tom Hiddleston, Sophie Okonedo, Rory Kinnear. Merlin told Kathy to buy it on Apple TV "or I can pirate it for you." Kenneth Branagh's Henry V (1989) -- "My number one movie that I recommend." "You don't even need to understand what they're saying. It'll still give you shivers." Mark Rylance: St. Crispin's Day speech at the Globe -- "It gives you a different kind of shivers, like a different part of your neck and your back." Merlin recited part of the speech from memory. The Death of Stalin (2017) -- "A very dark, very funny film" by Armando Iannucci. Veep / The Thick of It -- "It's gonna be difficult difficult lemon difficult." Both Iannucci. Led to Merlin imagining Matthew Butterick as a Veep restaurant reservation alias. Women Talking (2022) / Men (2022) -- Merlin's suggested double feature for mom's night. "Start with Women Talking, back with Men." Jessie Buckley, Rory Kinnear. Our Flag Means Death -- Merlin named his Mac Studio "Buttons" after Ewen Bremner's Mr. Buttons ("the guy from Trainspotting"). Rhys Darby, Kristian Nairn ("Hodor's on there. He's a big fella."). Fantastic Mr. Fox (2009) -- "Just to be available." Merlin's favorite line, from Mr. Kylie the possum wanting to know his job in the big plan. Music Vikingur Olafsson: Goldberg Variations (Deutsche Grammophon, 2023) -- Merlin's current obsession. "I care so intensely about that." Discovered after years of only knowing Glenn Gould. Glenn Gould: 1955 vs. 1981 Goldberg Variations -- The famous pair: 38 minutes of youthful showmanship vs. 51 minutes of deliberate structure. Public Enemy -- "Bring the Noise" -- Merlin rapped the full opening verse from memory when Kathy mentioned Flavor Flav. "Bass, how low can you go?" Poetry Gwendolyn Brooks -- "We Real Cool" (video of her 1983 Guggenheim reading) -- "We real cool. We jazz June. We die soon." Merlin on hearing poetry "in the air" vs. on the page. Sylvia Plath -- "Daddy" (her 1962 BBC recording) -- "You do not do, you do not do... you really hear something you didn't see on the page." Books & Podcasts Bessel van der Kolk on The Ezra Klein Show -- "One of my all-time favorite podcast episodes. It changed my life. Everything you know about trauma is screwing you up." Off Menu -- celebrities describe their dream meal. The Amanda Seyfried episode taught Merlin about a kind of olive he now puts on Brussels sprouts. Mr. Show with Bob and David -- source of the "hey everybody" drum bit Merlin does throughout. "I'm very, very, very specifically stealing it from a bit about the new Ku Klux Klan." Blank Check (Griffin Newman) -- source of "the great ___" bit. "I'll credit Griffin Newman for that bit." People James Thompson (PCalc, Dice by PCalc) -- "What if twenty-sided dice fell on your head?" Merlin on how James finds delight in close-to-the-metal Apple tech. Armando Iannucci -- "If you like English nerd comedy, he's really something." Simon Willison, Matt Webb, danah boyd -- people Merlin follows because "I'm just interested in your deal." Edgar Wright -- "I will just show up because I'm interested in what he's up to. I don't even care if I like his movie." Ecamm Live -- streaming app Kathy uses for her unicorn co-working sessions. Pre-Show (Superconductors only) LiDAR accessibility features on iPhone -- Merlin fiddled with it on the street, "pointing his phone at people for a very long time." Apple's breathing sleep LED -- the MacBook pulsing light. Kathy: "So relaxing, so unnecessary and delightful." Apple researched sleeping respiratory rates and chose the calmest end of the spectrum. Erich Brenn, plate spinner, on The Ed Sullivan Show -- the origin of "spinning plates" as a metaphor. 8 appearances in the 1950s-60s. Support Conduit with a Relay Membership

Thu, 26 Feb 2026 17:15:00 GMT http://relay.fm/conduit/122 http://relay.fm/conduit/122 There is No One True Anything with Merlin Mann 122 Kathy Campbell and Jay Miller Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. clean 5583 Jay is gone again, so Kathy brings back Merlin Mann to discuss productivity porn as well as a ton of other topics of import. Guest Starring: Merlin Mann Links and Show Notes: Checked Connections - Merlin ✅ - Working on collecting the old sites and Fives list - Kathy ✅ - Get ready for unicorning cowork Keep sending those MyConduit Connections to us on Discord and through Feedback! New Connections - Merlin - Keep working on the site thing - Kathy - Take things to the post office For Our Super Conductors: Pre-Show: LIDar on iOS. How do you know if you're ladder is against the right wall? Post-Show: Embracing the chaos Credits Music: When You Smile Executive Producers: Relay FM Discord Community Conduit e122 Links Merlin's One Good Things Where Everybody Knows Your Name: Judy Greer (Ted Danson, Conan O'Brien Network) -- "I went in thinking, oh, this looks really good, and I ended up liking it probably twice as much as I expected." Judy Greer -- Cheryl/Carol on Archer, Kitty Sanchez on Arrested Development. "It was neat to hear her talk about how important it was for her to get better at acting." Typora -- WYSIWYG Markdown editor ($15). "A really nice balance of what I'm looking for" -- discovered through the 5ives redesign work with Claude. Judi Dench speech on The Graham Norton Show -- "Made me cry." Kathy's One Good Thing Flavor Flav sponsoring the US women's hockey team -- Vegas celebration for the gold-medal team. Merlin responded by rapping "Bring the Noise" from memory. Merlin's Shows Do By Friday (with Alex Cox) Reconcilable Differences (with John Siracusa) Roderick on the Line (with John Roderick) Productivity / Publishing Inbox Zero -- "I'm the inbox zero guy." Merlin originated the concept; the world turned it into a marketing term. 43folders.com -- "In 2004, there were not a lot of websites about how to deal with your productivity problems as a Mac user." Back to Work (5by5) -- former podcast David Allen / Getting Things Done -- "He claims he's the laziest man in the world, and I've always admired that he says that." Danny O'Brien and the 2005 ETech "Life Hacks" talk -- "Danny and I are both so addled and odd and different... his energy was just incandescent to be around." The conference where Merlin's laptop had Wi-Fi for the first time. Site Meter -- "There's your life before site meter and your life after site meter." The little GIF badge that counted page loads and launched a million blog vanity spirals. 5ives & Typography 5ives -- Merlin's list site (2002), 450 lists, being revived. "I'm pleased with myself. I like that I made four hundred and fifty lists that some people thought were funny in the 2000s." Matthew Butterick -- fonts, Practical Typography. "One of those people where I'm just interested in your deal," like Simon Willison or Edgar Wright. Merlin bought the entire font set during a bout of situational depression and is finally using them for the 5ives redesign. Movies & TV The Hollow Crown (BBC) -- Trailer. "Look at that stacked cast." Ben Whishaw, Tom Hiddleston, Sophie Okonedo, Rory Kinnear. Merlin told Kathy to buy it on Apple TV "or I can pirate it for you." Kenneth Branagh's Henry V (1989) -- "My number one movie that I recommend." "You don't even need to understand what they're saying. It'll still give you shivers." Mark Rylance: St. Crispin's Day speech at the Globe -- "It gives you a different kind of shivers, like a different part of your neck and your back." Merlin recited part of the speech from memory. The Death of Stalin (2017) -- "A very dark, very funny film" by Armando Iannucci. Veep / The Thick of It -- "It's gonna be difficult difficult lemon difficult." Both Iannucci. Led to Merlin imagining Matthew Butterick as a Veep restaurant reservation alias. Women Talking (2022) / Men (2022) -- Merlin's suggested double feature for mom's night. "Start with Women Talking, back with Men." Jessie Buckley, Rory Kinnear. Our Flag Means Death -- Merlin named his Mac Studio "Buttons" after Ewen Bremner's Mr. Buttons ("the guy from Trainspotting"). Rhys Darby, Kristian Nairn ("Hodor's on there. He's a big fella."). Fantastic Mr. Fox (2009) -- "Just to be available." Merlin's favorite line, from Mr. Kylie the possum wanting to know his job in the big plan. Music Vikingur Olafsson: Goldberg Variations (Deutsche Grammophon, 2023) -- Merlin's current obsession. "I care so intensely about that." Discovered after years of only knowing Glenn Gould. Glenn Gould: 1955 vs. 1981 Goldberg Variations -- The famous pair: 38 minutes of youthful showmanship vs. 51 minutes of deliberate structure. Public Enemy -- "Bring the Noise" -- Merlin rapped the full opening verse from memory when Kathy mentioned Flavor Flav. "Bass, how low can you go?" Poetry Gwendolyn Brooks -- "We Real Cool" (video of her 1983 Guggenheim reading) -- "We real cool. We jazz June. We die soon." Merlin on hearing poetry "in the air" vs. on the page. Sylvia Plath -- "Daddy" (her 1962 BBC recording) -- "You do not do, you do not do... you really hear something you didn't see on the page." Books & Podcasts Bessel van der Kolk on The Ezra Klein Show -- "One of my all-time favorite podcast episodes. It changed my life. Everything you know about trauma is screwing you up." Off Menu -- celebrities describe their dream meal. The Amanda Seyfried episode taught Merlin about a kind of olive he now puts on Brussels sprouts. Mr. Show with Bob and David -- source of the "hey everybody" drum bit Merlin does throughout. "I'm very, very, very specifically stealing it from a bit about the new Ku Klux Klan." Blank Check (Griffin Newman) -- source of "the great ___" bit. "I'll credit Griffin Newman for that bit." People James Thompson (PCalc, Dice by PCalc) -- "What if twenty-sided dice fell on your head?" Merlin on how James finds delight in close-to-the-metal Apple tech. Armando Iannucci -- "If you like English nerd comedy, he's really something." Simon Willison, Matt Webb, danah boyd -- people Merlin follows because "I'm just interested in your deal." Edgar Wright -- "I will just show up because I'm interested in what he's up to. I don't even care if I like his movie." Ecamm Live -- streaming app Kathy uses for her unicorn co-working sessions. Pre-Show (Superconductors only) LiDAR accessibility features on iPhone -- Merlin fiddled with it on the street, "pointing his phone at people for a very long time." Apple's breathing sleep LED -- the MacBook pulsing light. Kathy: "So relaxing, so unnecessary and delightful." Apple researched sleeping respiratory rates and chose the calmest end of the spectrum. Erich Brenn, plate spinner, on The Ed Sullivan Show -- the origin of "spinning plates" as a metaphor. 8 appearances in the 1950s-60s. Support Conduit with a Relay Membership

Join Matt Johnson as he explores the intricacies of construction invoicing and strategic procurement at Triangle Construction Services. Discover insights into their expanding construction tech group, business model evolution, and the impact on lienholders and contractors. Matt discusses managing rework, change orders, and the role of LiDAR drone technology. Hear about fostering a positive company culture and embracing servant leadership.

Travelling as a blind passenger can be a mix of comedy, frustration, and tech discovery. Steven Scott shares his delayed journey from Vienna, encounters with chatty robots, inaccessible airports, and the reality of airport assistance for blind travellers—plus the apps and tools that saved the day. In this episode of Double Tap, Steven Scott returns from a snowy Vienna with stories of flight cancellations, unexpected overnight stays, and surreal encounters with plate-collecting robots. He dives into the challenges blind travellers face at airports, from being left at gates for hours to the lack of independence in exploring shops and cafés. Steven and Shaun discuss how apps like Curb to Car, Be My Eyes, and Aira made a huge difference, and debate the need for more flexible airport assistance—perhaps even an “upgrade” option for travellers who value autonomy. The conversation also explores future solutions, including AI-powered navigation, LIDAR apps like EyeGuide, and why better Wi-Fi could transform independent travel experiences. Relevant Links Be My Eyes: https://www.bemyeyes.com Aira: https://aira.io Curb to Car on iOS: https://apps.apple.com/app/curb-to-car/id6475301159 I Guide App: https://apps.apple.com/app/i-guide/id6734567890 Find Double Tap online: YouTube, Double Tap Website---Follow on:YouTube: https://www.doubletaponair.com/youtubeX (formerly Twitter): https://www.doubletaponair.com/xInstagram: https://www.doubletaponair.com/instagramTikTok: https://www.doubletaponair.com/tiktokThreads: https://www.doubletaponair.com/threadsFacebook: https://www.doubletaponair.com/facebookLinkedIn: https://www.doubletaponair.com/linkedin Subscribe to the Podcast:Apple: https://www.doubletaponair.com/appleSpotify: https://www.doubletaponair.com/spotifyRSS: https://www.doubletaponair.com/podcastiHeadRadio: https://www.doubletaponair.com/iheart About Double TapHosted by the insightful duo, Steven Scott and Shaun Preece, Double Tap is a treasure trove of information for anyone who's blind or partially sighted and has a passion for tech. Steven and Shaun not only demystify tech, but they also regularly feature interviews and welcome guests from the community, fostering an interactive and engaging environment. Tune in every day of the week, and you'll discover how technology can seamlessly integrate into your life, enhancing daily tasks and experiences, even if your sight is limited. "Double Tap" is a registered trademark of Double Tap Productions Inc. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Want AI that works the first time instead of the tenth? We sit down with Andreas Schachl, co-founder of Responsible Annotation Services, to unpack the quiet truth behind reliable models: ethical, high-quality training data produced by people who take clarity and precision seriously. Andreas shares how a single internship sparked a company built around neurodivergent talent, turning data labeling from a churn task into a strategic advantage.We walk through why annotation isn't going anywhere, even with foundation models and smarter tools. When you're training on private, business-owned data across text, images, audio, video, and LiDAR, you need a human in the loop and documentation you can defend. Andreas explains how his team co-authors rigorous annotation handbooks with clients, translating fuzzy goals into exact rules, edge cases, and review procedures. The payoff is real: higher consistency, fewer iterations, and a clear compliance trail for regulators and auditors.Bias mitigation becomes a practice, not a promise. A neurodivergent lens exposes hidden assumptions and pushes for instructions that are unambiguous and testable. We explore practical systems—daily stand-ups, structured chat, and even “coffee calls” with agendas—that help people do their best focused work. We also confront the ethics of the global annotation supply chain and outline a different path: EU contracts, fair wages, social worker support, and leadership that values diligence over hype. From 2D images to complex 3D point clouds, we show how modern tooling plus human judgment builds AI you can trust.If you care about responsible AI, data quality, and making models perform sooner with less guesswork, this conversation is your blueprint. Subscribe, share with a colleague wrestling with training data, and leave a review with your biggest annotation challenge—we'll tackle it in a future episode.Send a textSupport the showFollow axschat on social media.Bluesky:Antonio https://bsky.app/profile/akwyz.com Debra https://bsky.app/profile/debraruh.bsky.social Neil https://bsky.app/profile/neilmilliken.bsky.social axschat https://bsky.app/profile/axschat.bsky.social LinkedInhttps://www.linkedin.com/in/antoniovieirasantos/ https://www.linkedin.com/company/axschat/ https://www.linkedin.com/in/neilmilliken/Vimeohttps://vimeo.com/akwyzhttps://twitter.com/axschathttps://twitter.com/AkwyZhttps://twitter.com/neilmillikenhttps://twitter.com/debraruh

What does early 2026 mean for the American drone industry?In this episode of Your Drone Questions. Answered, we sit down with WISPR Systems to talk about their journey from a Mississippi startup to a nationally recognized U.S. drone manufacturer — and how recent regulatory changes are impacting the entire UAS landscape.John McArthur shares how WISPR Systems evolved from building drones for professors at Mississippi State University to becoming a major player in the surveying and mapping space. You'll hear how the company made a strategic decision to “plant their flag” in one vertical, master complex technologies like RTK, PPK, LiDAR, and photogrammetry, and build partnerships with key industry leaders.We also break down:The December 2025 Federal Highway and FCC announcementsWhat NDAA compliance actually means (and why it's often misunderstood)The difference between American-made, NDA compliant, Green List, and Blue List dronesHow supply chain realities — especially batteries and payload components — are shaping U.S. manufacturingWhat this all means for contractors, DOT projects, and commercial operatorsWhether waivers may provide clarity for allied-country payloads like Sony camerasIf you've been confused about compliance language, worried about what you can legally fly, or wondering how U.S. drone manufacturing is evolving under new federal priorities — this conversation brings clarity.You'll also learn about WISPR Ranger Pro and the SkyScout series — including how they positioned their platform as a compact, open-payload alternative for professional surveying and mapping teams transitioning from DJI ecosystems.For more information or to connect with the Whisper team, visit: 

The comparison between Tesla's vision-only approach and Waymo's use of LIDAR highlights a fundamental disagreement in self-driving philosophy. Tesla relies exclusively on visual cameras, while Waymo utilizes LIDAR (Light Detection and Ranging) as a primary sensor to map the vehicle's surroundings.The sources provide the following insights into how these two systems compare:Technical Philosophy and Sensor Suite• Tesla (Vision-Only): Tesla's strategy is based on the belief that vision is the only necessary input for self-driving, similar to how the human nervous system functions. However, critics in the sources argue that Tesla has "blown what could have been a data advantage" by refusing to use additional sensors like LIDAR.• Waymo (LIDAR-based): Waymo's system is often viewed as "far superior" in its current state because LIDAR provides precise depth and spatial data that cameras alone may struggle to replicate.Safety and Performance Records• Crash Rates: Reports indicate that Tesla's robotaxis have a crash rate approximately four times higher than human drivers, based on data from Austin where the fleet logged four crashes in four months. Conversely, some users suggest that Waymo operates with fewer accidents than human drivers.• Reliability: User experiences with Waymo are frequently described as "almost flawless" or working "pretty flawlessly" in cities like San Francisco and Austin. In contrast, Tesla's system is described by some as "lagging on roads" and currently under investigation for incidents, such as those involving railroads.Current Limitations• Waymo's Weaknesses: Despite its perceived superiority, Waymo still faces challenges. Users have noted that the vehicles can struggle in heavy rain or become confused by temporary road closures for events. Additionally, some reports suggest Waymo may rely on remote operators in other countries to assist the vehicles.• Tesla's Weaknesses: Critics argue that it is impossible to compete with LIDAR using only visual cameras. Further, there are reports that Tesla's "driverless" tests still involve human safety monitors following the robotaxis in trailing cars.The Debate on "Vision-Only"While some argue that a vision-only system will "never ever" be as good as LIDAR, others suggest that technology may eventually advance to a point where vision is sufficient. However, the current consensus among the provided sources is that LIDAR provides a level of safety and reliability that Tesla's camera-based system has yet to achieve

The future isn't just AI that thinks—it's AI that SEES and INTERACTS with the physical world.Simon Erickson chats with Emmett Savage (MyWallStreet & Prophet founder) to break down Ouster (OUST)—the company making "seeing eyes for AI" through breakthrough solid-state LIDAR technology. No moving parts. Pure semiconductor engineering. And it's already deployed in over 100,000 sensors.This isn't a paved road—it's early and risky. But we might be looking at one of the ultimate building blocks of seeing machines.Stocks Discussed:Ouster (OUST) - Featured stockVertical Aerospace - Previous EVTOL discussionServ Robotics - Delivery robotsTesla, Apple - Tier-1 customersiRobot, Mobileye, InvenSense - Historical comparisonsNext Episode Monday (Feb 2): Simon reveals the space where his NEXT 7investing recommendation operates (Groundhog Day special!)Next Episode Wednesday (Feb 4): Emmett returns with a THIRD off-radar stock pick

From rewriting Google's search stack in the early 2000s to reviving sparse trillion-parameter models and co-designing TPUs with frontier ML research, Jeff Dean has quietly shaped nearly every layer of the modern AI stack. As Chief AI Scientist at Google and a driving force behind Gemini, Jeff has lived through multiple scaling revolutions from CPUs and sharded indices to multimodal models that reason across text, video, and code.Jeff joins us to unpack what it really means to “own the Pareto frontier,” why distillation is the engine behind every Flash model breakthrough, how energy (in picojoules) not FLOPs is becoming the true bottleneck, what it was like leading the charge to unify all of Google's AI teams, and why the next leap won't come from bigger context windows alone, but from systems that give the illusion of attending to trillions of tokens.We discuss:* Jeff's early neural net thesis in 1990: parallel training before it was cool, why he believed scaling would win decades early, and the “bigger model, more data, better results” mantra that held for 15 years* The evolution of Google Search: sharding, moving the entire index into memory in 2001, softening query semantics pre-LLMs, and why retrieval pipelines already resemble modern LLM systems* Pareto frontier strategy: why you need both frontier “Pro” models and low-latency “Flash” models, and how distillation lets smaller models surpass prior generations* Distillation deep dive: ensembles → compression → logits as soft supervision, and why you need the biggest model to make the smallest one good* Latency as a first-class objective: why 10–50x lower latency changes UX entirely, and how future reasoning workloads will demand 10,000 tokens/sec* Energy-based thinking: picojoules per bit, why moving data costs 1000x more than a multiply, batching through the lens of energy, and speculative decoding as amortization* TPU co-design: predicting ML workloads 2–6 years out, speculative hardware features, precision reduction, sparsity, and the constant feedback loop between model architecture and silicon* Sparse models and “outrageously large” networks: trillions of parameters with 1–5% activation, and why sparsity was always the right abstraction* Unified vs. specialized models: abandoning symbolic systems, why general multimodal models tend to dominate vertical silos, and when vertical fine-tuning still makes sense* Long context and the illusion of scale: beyond needle-in-a-haystack benchmarks toward systems that narrow trillions of tokens to 117 relevant documents* Personalized AI: attending to your emails, photos, and documents (with permission), and why retrieval + reasoning will unlock deeply personal assistants* Coding agents: 50 AI interns, crisp specifications as a new core skill, and how ultra-low latency will reshape human–agent collaboration* Why ideas still matter: transformers, sparsity, RL, hardware, systems — scaling wasn't blind; the pieces had to multiply togetherShow Notes:* Gemma 3 Paper* Gemma 3* Gemini 2.5 Report* Jeff Dean's “Software Engineering Advice fromBuilding Large-Scale Distributed Systems” Presentation (with Back of the Envelope Calculations)* Latency Numbers Every Programmer Should Know by Jeff Dean* The Jeff Dean Facts* Jeff Dean Google Bio* Jeff Dean on “Important AI Trends” @Stanford AI Club* Jeff Dean & Noam Shazeer — 25 years at Google (Dwarkesh)—Jeff Dean* LinkedIn: https://www.linkedin.com/in/jeff-dean-8b212555* X: https://x.com/jeffdeanGoogle* https://google.com* https://deepmind.googleFull Video EpisodeTimestamps00:00:04 — Introduction: Alessio & Swyx welcome Jeff Dean, chief AI scientist at Google, to the Latent Space podcast00:00:30 — Owning the Pareto Frontier & balancing frontier vs low-latency models00:01:31 — Frontier models vs Flash models + role of distillation00:03:52 — History of distillation and its original motivation00:05:09 — Distillation's role in modern model scaling00:07:02 — Model hierarchy (Flash, Pro, Ultra) and distillation sources00:07:46 — Flash model economics & wide deployment00:08:10 — Latency importance for complex tasks00:09:19 — Saturation of some tasks and future frontier tasks00:11:26 — On benchmarks, public vs internal00:12:53 — Example long-context benchmarks & limitations00:15:01 — Long-context goals: attending to trillions of tokens00:16:26 — Realistic use cases beyond pure language00:18:04 — Multimodal reasoning and non-text modalities00:19:05 — Importance of vision & motion modalities00:20:11 — Video understanding example (extracting structured info)00:20:47 — Search ranking analogy for LLM retrieval00:23:08 — LLM representations vs keyword search00:24:06 — Early Google search evolution & in-memory index00:26:47 — Design principles for scalable systems00:28:55 — Real-time index updates & recrawl strategies00:30:06 — Classic “Latency numbers every programmer should know”00:32:09 — Cost of memory vs compute and energy emphasis00:34:33 — TPUs & hardware trade-offs for serving models00:35:57 — TPU design decisions & co-design with ML00:38:06 — Adapting model architecture to hardware00:39:50 — Alternatives: energy-based models, speculative decoding00:42:21 — Open research directions: complex workflows, RL00:44:56 — Non-verifiable RL domains & model evaluation00:46:13 — Transition away from symbolic systems toward unified LLMs00:47:59 — Unified models vs specialized ones00:50:38 — Knowledge vs reasoning & retrieval + reasoning00:52:24 — Vertical model specialization & modules00:55:21 — Token count considerations for vertical domains00:56:09 — Low resource languages & contextual learning00:59:22 — Origins: Dean's early neural network work01:10:07 — AI for coding & human–model interaction styles01:15:52 — Importance of crisp specification for coding agents01:19:23 — Prediction: personalized models & state retrieval01:22:36 — Token-per-second targets (10k+) and reasoning throughput01:23:20 — Episode conclusion and thanksTranscriptAlessio Fanelli [00:00:04]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, founder of Kernel Labs, and I'm joined by Swyx, editor of Latent Space. Shawn Wang [00:00:11]: Hello, hello. We're here in the studio with Jeff Dean, chief AI scientist at Google. Welcome. Thanks for having me. It's a bit surreal to have you in the studio. I've watched so many of your talks, and obviously your career has been super legendary. So, I mean, congrats. I think the first thing must be said, congrats on owning the Pareto Frontier.Jeff Dean [00:00:30]: Thank you, thank you. Pareto Frontiers are good. It's good to be out there.Shawn Wang [00:00:34]: Yeah, I mean, I think it's a combination of both. You have to own the Pareto Frontier. You have to have like frontier capability, but also efficiency, and then offer that range of models that people like to use. And, you know, some part of this was started because of your hardware work. Some part of that is your model work, and I'm sure there's lots of secret sauce that you guys have worked on cumulatively. But, like, it's really impressive to see it all come together in, like, this slittily advanced.Jeff Dean [00:01:04]: Yeah, yeah. I mean, I think, as you say, it's not just one thing. It's like a whole bunch of things up and down the stack. And, you know, all of those really combine to help make UNOS able to make highly capable large models, as well as, you know, software techniques to get those large model capabilities into much smaller, lighter weight models that are, you know, much more cost effective and lower latency, but still, you know, quite capable for their size. Yeah.Alessio Fanelli [00:01:31]: How much pressure do you have on, like, having the lower bound of the Pareto Frontier, too? I think, like, the new labs are always trying to push the top performance frontier because they need to raise more money and all of that. And you guys have billions of users. And I think initially when you worked on the CPU, you were thinking about, you know, if everybody that used Google, we use the voice model for, like, three minutes a day, they were like, you need to double your CPU number. Like, what's that discussion today at Google? Like, how do you prioritize frontier versus, like, we have to do this? How do we actually need to deploy it if we build it?Jeff Dean [00:02:03]: Yeah, I mean, I think we always want to have models that are at the frontier or pushing the frontier because I think that's where you see what capabilities now exist that didn't exist at the sort of slightly less capable last year's version or last six months ago version. At the same time, you know, we know those are going to be really useful for a bunch of use cases, but they're going to be a bit slower and a bit more expensive than people might like for a bunch of other broader models. So I think what we want to do is always have kind of a highly capable sort of affordable model that enables a whole bunch of, you know, lower latency use cases. People can use them for agentic coding much more readily and then have the high-end, you know, frontier model that is really useful for, you know, deep reasoning, you know, solving really complicated math problems, those kinds of things. And it's not that. One or the other is useful. They're both useful. So I think we'd like to do both. And also, you know, through distillation, which is a key technique for making the smaller models more capable, you know, you have to have the frontier model in order to then distill it into your smaller model. So it's not like an either or choice. You sort of need that in order to actually get a highly capable, more modest size model. Yeah.Alessio Fanelli [00:03:24]: I mean, you and Jeffrey came up with the solution in 2014.Jeff Dean [00:03:28]: Don't forget, L'Oreal Vinyls as well. Yeah, yeah.Alessio Fanelli [00:03:30]: A long time ago. But like, I'm curious how you think about the cycle of these ideas, even like, you know, sparse models and, you know, how do you reevaluate them? How do you think about in the next generation of model, what is worth revisiting? Like, yeah, they're just kind of like, you know, you worked on so many ideas that end up being influential, but like in the moment, they might not feel that way necessarily. Yeah.Jeff Dean [00:03:52]: I mean, I think distillation was originally motivated because we were seeing that we had a very large image data set at the time, you know, 300 million images that we could train on. And we were seeing that if you create specialists for different subsets of those image categories, you know, this one's going to be really good at sort of mammals, and this one's going to be really good at sort of indoor room scenes or whatever, and you can cluster those categories and train on an enriched stream of data after you do pre-training on a much broader set of images. You get much better performance. If you then treat that whole set of maybe 50 models you've trained as a large ensemble, but that's not a very practical thing to serve, right? So distillation really came about from the idea of, okay, what if we want to actually serve that and train all these independent sort of expert models and then squish it into something that actually fits in a form factor that you can actually serve? And that's, you know, not that different from what we're doing today. You know, often today we're instead of having an ensemble of 50 models. We're having a much larger scale model that we then distill into a much smaller scale model.Shawn Wang [00:05:09]: Yeah. A part of me also wonders if distillation also has a story with the RL revolution. So let me maybe try to articulate what I mean by that, which is you can, RL basically spikes models in a certain part of the distribution. And then you have to sort of, well, you can spike models, but usually sometimes... It might be lossy in other areas and it's kind of like an uneven technique, but you can probably distill it back and you can, I think that the sort of general dream is to be able to advance capabilities without regressing on anything else. And I think like that, that whole capability merging without loss, I feel like it's like, you know, some part of that should be a distillation process, but I can't quite articulate it. I haven't seen much papers about it.Jeff Dean [00:06:01]: Yeah, I mean, I tend to think of one of the key advantages of distillation is that you can have a much smaller model and you can have a very large, you know, training data set and you can get utility out of making many passes over that data set because you're now getting the logits from the much larger model in order to sort of coax the right behavior out of the smaller model that you wouldn't otherwise get with just the hard labels. And so, you know, I think that's what we've observed. Is you can get, you know, very close to your largest model performance with distillation approaches. And that seems to be, you know, a nice sweet spot for a lot of people because it enables us to kind of, for multiple Gemini generations now, we've been able to make the sort of flash version of the next generation as good or even substantially better than the previous generations pro. And I think we're going to keep trying to do that because that seems like a good trend to follow.Shawn Wang [00:07:02]: So, Dara asked, so it was the original map was Flash Pro and Ultra. Are you just sitting on Ultra and distilling from that? Is that like the mother load?Jeff Dean [00:07:12]: I mean, we have a lot of different kinds of models. Some are internal ones that are not necessarily meant to be released or served. Some are, you know, our pro scale model and we can distill from that as well into our Flash scale model. So I think, you know, it's an important set of capabilities to have and also inference time scaling. It can also be a useful thing to improve the capabilities of the model.Shawn Wang [00:07:35]: And yeah, yeah, cool. Yeah. And obviously, I think the economy of Flash is what led to the total dominance. I think the latest number is like 50 trillion tokens. I don't know. I mean, obviously, it's changing every day.Jeff Dean [00:07:46]: Yeah, yeah. But, you know, by market share, hopefully up.Shawn Wang [00:07:50]: No, I mean, there's no I mean, there's just the economics wise, like because Flash is so economical, like you can use it for everything. Like it's in Gmail now. It's in YouTube. Like it's yeah. It's in everything.Jeff Dean [00:08:02]: We're using it more in our search products of various AI mode reviews.Shawn Wang [00:08:05]: Oh, my God. Flash past the AI mode. Oh, my God. Yeah, that's yeah, I didn't even think about that.Jeff Dean [00:08:10]: I mean, I think one of the things that is quite nice about the Flash model is not only is it more affordable, it's also a lower latency. And I think latency is actually a pretty important characteristic for these models because we're going to want models to do much more complicated things that are going to involve, you know, generating many more tokens from when you ask the model to do so. So, you know, if you're going to ask the model to do something until it actually finishes what you ask it to do, because you're going to ask now, not just write me a for loop, but like write me a whole software package to do X or Y or Z. And so having low latency systems that can do that seems really important. And Flash is one direction, one way of doing that. You know, obviously our hardware platforms enable a bunch of interesting aspects of our, you know, serving stack as well, like TPUs, the interconnect between. Chips on the TPUs is actually quite, quite high performance and quite amenable to, for example, long context kind of attention operations, you know, having sparse models with lots of experts. These kinds of things really, really matter a lot in terms of how do you make them servable at scale.Alessio Fanelli [00:09:19]: Yeah. Does it feel like there's some breaking point for like the proto Flash distillation, kind of like one generation delayed? I almost think about almost like the capability as a. In certain tasks, like the pro model today is a saturated, some sort of task. So next generation, that same task will be saturated at the Flash price point. And I think for most of the things that people use models for at some point, the Flash model in two generation will be able to do basically everything. And how do you make it economical to like keep pushing the pro frontier when a lot of the population will be okay with the Flash model? I'm curious how you think about that.Jeff Dean [00:09:59]: I mean, I think that's true. If your distribution of what people are asking people, the models to do is stationary, right? But I think what often happens is as the models become more capable, people ask them to do more, right? So, I mean, I think this happens in my own usage. Like I used to try our models a year ago for some sort of coding task, and it was okay at some simpler things, but wouldn't do work very well for more complicated things. And since then, we've improved dramatically on the more complicated coding tasks. And now I'll ask it to do much more complicated things. And I think that's true, not just of coding, but of, you know, now, you know, can you analyze all the, you know, renewable energy deployments in the world and give me a report on solar panel deployment or whatever. That's a very complicated, you know, more complicated task than people would have asked a year ago. And so you are going to want more capable models to push the frontier in the absence of what people ask the models to do. And that also then gives us. Insight into, okay, where does the, where do things break down? How can we improve the model in these, these particular areas, uh, in order to sort of, um, make the next generation even better.Alessio Fanelli [00:11:11]: Yeah. Are there any benchmarks or like test sets they use internally? Because it's almost like the same benchmarks get reported every time. And it's like, all right, it's like 99 instead of 97. Like, how do you have to keep pushing the team internally to it? Or like, this is what we're building towards. Yeah.Jeff Dean [00:11:26]: I mean, I think. Benchmarks, particularly external ones that are publicly available. Have their utility, but they often kind of have a lifespan of utility where they're introduced and maybe they're quite hard for current models. You know, I, I like to think of the best kinds of benchmarks are ones where the initial scores are like 10 to 20 or 30%, maybe, but not higher. And then you can sort of work on improving that capability for, uh, whatever it is, the benchmark is trying to assess and get it up to like 80, 90%, whatever. I, I think once it hits kind of 95% or something, you get very diminishing returns from really focusing on that benchmark, cuz it's sort of, it's either the case that you've now achieved that capability, or there's also the issue of leakage in public data or very related kind of data being, being in your training data. Um, so we have a bunch of held out internal benchmarks that we really look at where we know that wasn't represented in the training data at all. There are capabilities that we want the model to have. Um, yeah. Yeah. Um, that it doesn't have now, and then we can work on, you know, assessing, you know, how do we make the model better at these kinds of things? Is it, we need different kind of data to train on that's more specialized for this particular kind of task. Do we need, um, you know, a bunch of, uh, you know, architectural improvements or some sort of, uh, model capability improvements, you know, what would help make that better?Shawn Wang [00:12:53]: Is there, is there such an example that you, uh, a benchmark inspired in architectural improvement? Like, uh, I'm just kind of. Jumping on that because you just.Jeff Dean [00:13:02]: Uh, I mean, I think some of the long context capability of the, of the Gemini models that came, I guess, first in 1.5 really were about looking at, okay, we want to have, um, you know,Shawn Wang [00:13:15]: immediately everyone jumped to like completely green charts of like, everyone had, I was like, how did everyone crack this at the same time? Right. Yeah. Yeah.Jeff Dean [00:13:23]: I mean, I think, um, and once you're set, I mean, as you say that needed single needle and a half. Hey, stack benchmark is really saturated for at least context links up to 1, 2 and K or something. Don't actually have, you know, much larger than 1, 2 and 8 K these days or two or something. We're trying to push the frontier of 1 million or 2 million context, which is good because I think there are a lot of use cases where. Yeah. You know, putting a thousand pages of text or putting, you know, multiple hour long videos and the context and then actually being able to make use of that as useful. Try to, to explore the über graduation are fairly large. But the single needle in a haystack benchmark is sort of saturated. So you really want more complicated, sort of multi-needle or more realistic, take all this content and produce this kind of answer from a long context that sort of better assesses what it is people really want to do with long context. Which is not just, you know, can you tell me the product number for this particular thing?Shawn Wang [00:14:31]: Yeah, it's retrieval. It's retrieval within machine learning. It's interesting because I think the more meta level I'm trying to operate at here is you have a benchmark. You're like, okay, I see the architectural thing I need to do in order to go fix that. But should you do it? Because sometimes that's an inductive bias, basically. It's what Jason Wei, who used to work at Google, would say. Exactly the kind of thing. Yeah, you're going to win. Short term. Longer term, I don't know if that's going to scale. You might have to undo that.Jeff Dean [00:15:01]: I mean, I like to sort of not focus on exactly what solution we're going to derive, but what capability would you want? And I think we're very convinced that, you know, long context is useful, but it's way too short today. Right? Like, I think what you would really want is, can I attend to the internet while I answer my question? Right? But that's not going to happen. I think that's going to be solved by purely scaling the existing solutions, which are quadratic. So a million tokens kind of pushes what you can do. You're not going to do that to a trillion tokens, let alone, you know, a billion tokens, let alone a trillion. But I think if you could give the illusion that you can attend to trillions of tokens, that would be amazing. You'd find all kinds of uses for that. You would have attend to the internet. You could attend to the pixels of YouTube and the sort of deeper representations that we can find. You could attend to the form for a single video, but across many videos, you know, on a personal Gemini level, you could attend to all of your personal state with your permission. So like your emails, your photos, your docs, your plane tickets you have. I think that would be really, really useful. And the question is, how do you get algorithmic improvements and system level improvements that get you to something where you actually can attend to trillions of tokens? Right. In a meaningful way. Yeah.Shawn Wang [00:16:26]: But by the way, I think I did some math and it's like, if you spoke all day, every day for eight hours a day, you only generate a maximum of like a hundred K tokens, which like very comfortably fits.Jeff Dean [00:16:38]: Right. But if you then say, okay, I want to be able to understand everything people are putting on videos.Shawn Wang [00:16:46]: Well, also, I think that the classic example is you start going beyond language into like proteins and whatever else is extremely information dense. Yeah. Yeah.Jeff Dean [00:16:55]: I mean, I think one of the things about Gemini's multimodal aspects is we've always wanted it to be multimodal from the start. And so, you know, that sometimes to people means text and images and video sort of human-like and audio, audio, human-like modalities. But I think it's also really useful to have Gemini know about non-human modalities. Yeah. Like LIDAR sensor data from. Yes. Say, Waymo vehicles or. Like robots or, you know, various kinds of health modalities, x-rays and MRIs and imaging and genomics information. And I think there's probably hundreds of modalities of data where you'd like the model to be able to at least be exposed to the fact that this is an interesting modality and has certain meaning in the world. Where even if you haven't trained on all the LIDAR data or MRI data, you could have, because maybe that's not, you know, it doesn't make sense in terms of trade-offs of. You know, what you include in your main pre-training data mix, at least including a little bit of it is actually quite useful. Yeah. Because it sort of tempts the model that this is a thing.Shawn Wang [00:18:04]: Yeah. Do you believe, I mean, since we're on this topic and something I just get to ask you all the questions I always wanted to ask, which is fantastic. Like, are there some king modalities, like modalities that supersede all the other modalities? So a simple example was Vision can, on a pixel level, encode text. And DeepSeq had this DeepSeq CR paper that did that. Vision. And Vision has also been shown to maybe incorporate audio because you can do audio spectrograms and that's, that's also like a Vision capable thing. Like, so, so maybe Vision is just the king modality and like. Yeah.Jeff Dean [00:18:36]: I mean, Vision and Motion are quite important things, right? Motion. Well, like video as opposed to static images, because I mean, there's a reason evolution has evolved eyes like 23 independent ways, because it's such a useful capability for sensing the world around you, which is really what we want these models to be. So I think the only thing that we can be able to do is interpret the things we're seeing or the things we're paying attention to and then help us in using that information to do things. Yeah.Shawn Wang [00:19:05]: I think motion, you know, I still want to shout out, I think Gemini, still the only native video understanding model that's out there. So I use it for YouTube all the time. Nice.Jeff Dean [00:19:15]: Yeah. Yeah. I mean, it's actually, I think people kind of are not necessarily aware of what the Gemini models can actually do. Yeah. Like I have an example I've used in one of my talks. It had like, it was like a YouTube highlight video of 18 memorable sports moments across the last 20 years or something. So it has like Michael Jordan hitting some jump shot at the end of the finals and, you know, some soccer goals and things like that. And you can literally just give it the video and say, can you please make me a table of what all these different events are? What when the date is when they happened? And a short description. And so you get like now an 18 row table of that information extracted from the video, which is, you know, not something most people think of as like a turn video into sequel like table.Alessio Fanelli [00:20:11]: Has there been any discussion inside of Google of like, you mentioned tending to the whole internet, right? Google, it's almost built because a human cannot tend to the whole internet and you need some sort of ranking to find what you need. Yep. That ranking is like much different for an LLM because you can expect a person to look at maybe the first five, six links in a Google search versus for an LLM. Should you expect to have 20 links that are highly relevant? Like how do you internally figure out, you know, how do we build the AI mode that is like maybe like much broader search and span versus like the more human one? Yeah.Jeff Dean [00:20:47]: I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. With a giant number of web pages in our index, many of them are not relevant. So you identify a subset of them that are relevant with very lightweight kinds of methods. You know, you're down to like 30,000 documents or something. And then you gradually refine that to apply more and more sophisticated algorithms and more and more sophisticated sort of signals of various kinds in order to get down to ultimately what you show, which is, you know, the final 10 results or, you know, 10 results plus. Other kinds of information. And I think an LLM based system is not going to be that dissimilar, right? You're going to attend to trillions of tokens, but you're going to want to identify, you know, what are the 30,000 ish documents that are with the, you know, maybe 30 million interesting tokens. And then how do you go from that into what are the 117 documents I really should be paying attention to in order to carry out the tasks that the user has asked? And I think, you know, you can imagine systems where you have, you know, a lot of highly parallel processing to identify those initial 30,000 candidates, maybe with very lightweight kinds of models. Then you have some system that sort of helps you narrow down from 30,000 to the 117 with maybe a little bit more sophisticated model or set of models. And then maybe the final model is the thing that looks. So the 117 things that might be your most capable model. So I think it has to, it's going to be some system like that, that is really enables you to give the illusion of attending to trillions of tokens. Sort of the way Google search gives you, you know, not the illusion, but you are searching the internet, but you're finding, you know, a very small subset of things that are, that are relevant.Shawn Wang [00:22:47]: Yeah. I often tell a lot of people that are not steeped in like Google search history that, well, you know, like Bert was. Like he was like basically immediately inside of Google search and that improves results a lot, right? Like I don't, I don't have any numbers off the top of my head, but like, I'm sure you guys, that's obviously the most important numbers to Google. Yeah.Jeff Dean [00:23:08]: I mean, I think going to an LLM based representation of text and words and so on enables you to get out of the explicit hard notion of, of particular words having to be on the page, but really getting at the notion of this topic of this page or this page. Paragraph is highly relevant to this query. Yeah.Shawn Wang [00:23:28]: I don't think people understand how much LLMs have taken over all these very high traffic system, very high traffic. Yeah. Like it's Google, it's YouTube. YouTube has this like semantics ID thing where it's just like every token or every item in the vocab is a YouTube video or something that predicts the video using a code book, which is absurd to me for YouTube size.Jeff Dean [00:23:50]: And then most recently GROK also for, for XAI, which is like, yeah. I mean, I'll call out even before LLMs were used extensively in search, we put a lot of emphasis on softening the notion of what the user actually entered into the query.Shawn Wang [00:24:06]: So do you have like a history of like, what's the progression? Oh yeah.Jeff Dean [00:24:09]: I mean, I actually gave a talk in, uh, I guess, uh, web search and data mining conference in 2009, uh, where we never actually published any papers about the origins of Google search, uh, sort of, but we went through sort of four or five or six. generations, four or five or six generations of, uh, redesigning of the search and retrieval system, uh, from about 1999 through 2004 or five. And that talk is really about that evolution. And one of the things that really happened in 2001 was we were sort of working to scale the system in multiple dimensions. So one is we wanted to make our index bigger, so we could retrieve from a larger index, which always helps your quality in general. Uh, because if you don't have the page in your index, you're going to not do well. Um, and then we also needed to scale our capacity because we were, our traffic was growing quite extensively. Um, and so we had, you know, a sharded system where you have more and more shards as the index grows, you have like 30 shards. And then if you want to double the index size, you make 60 shards so that you can bound the latency by which you respond for any particular user query. Um, and then as traffic grows, you add, you add more and more replicas of each of those. And so we eventually did the math that realized that in a data center where we had say 60 shards and, um, you know, 20 copies of each shard, we now had 1200 machines, uh, with disks. And we did the math and we're like, Hey, one copy of that index would actually fit in memory across 1200 machines. So in 2001, we introduced, uh, we put our entire index in memory and what that enabled from a quality perspective was amazing. Um, and so we had more and more replicas of each of those. Before you had to be really careful about, you know, how many different terms you looked at for a query, because every one of them would involve a disk seek on every one of the 60 shards. And so you, as you make your index bigger, that becomes even more inefficient. But once you have the whole index in memory, it's totally fine to have 50 terms you throw into the query from the user's original three or four word query, because now you can add synonyms like restaurant and restaurants and cafe and, uh, you know, things like that. Uh, bistro and all these things. And you can suddenly start, uh, sort of really, uh, getting at the meaning of the word as opposed to the exact semantic form the user typed in. And that was, you know, 2001, very much pre LLM, but really it was about softening the, the strict definition of what the user typed in order to get at the meaning.Alessio Fanelli [00:26:47]: What are like principles that you use to like design the systems, especially when you have, I mean, in 2001, the internet is like. Doubling, tripling every year in size is not like, uh, you know, and I think today you kind of see that with LLMs too, where like every year the jumps in size and like capabilities are just so big. Are there just any, you know, principles that you use to like, think about this? Yeah.Jeff Dean [00:27:08]: I mean, I think, uh, you know, first, whenever you're designing a system, you want to understand what are the sort of design parameters that are going to be most important in designing that, you know? So, you know, how many queries per second do you need to handle? How big is the internet? How big is the index you need to handle? How much data do you need to keep for every document in the index? How are you going to look at it when you retrieve things? Um, what happens if traffic were to double or triple, you know, will that system work well? And I think a good design principle is you're going to want to design a system so that the most important characteristics could scale by like factors of five or 10, but probably not beyond that because often what happens is if you design a system for X. And something suddenly becomes a hundred X, that would enable a very different point in the design space that would not make sense at X. But all of a sudden at a hundred X makes total sense. So like going from a disk space index to a in memory index makes a lot of sense once you have enough traffic, because now you have enough replicas of the sort of state on disk that those machines now actually can hold, uh, you know, a full copy of the, uh, index and memory. Yeah. And that all of a sudden enabled. A completely different design that wouldn't have been practical before. Yeah. Um, so I'm, I'm a big fan of thinking through designs in your head, just kind of playing with the design space a little before you actually do a lot of writing of code. But, you know, as you said, in the early days of Google, we were growing the index, uh, quite extensively. We were growing the update rate of the index. So the update rate actually is the parameter that changed the most. Surprising. So it used to be once a month.Shawn Wang [00:28:55]: Yeah.Jeff Dean [00:28:56]: And then we went to a system that could update any particular page in like sub one minute. Okay.Shawn Wang [00:29:02]: Yeah. Because this is a competitive advantage, right?Jeff Dean [00:29:04]: Because all of a sudden news related queries, you know, if you're, if you've got last month's news index, it's not actually that useful for.Shawn Wang [00:29:11]: News is a special beast. Was there any, like you could have split it onto a separate system.Jeff Dean [00:29:15]: Well, we did. We launched a Google news product, but you also want news related queries that people type into the main index to also be sort of updated.Shawn Wang [00:29:23]: So, yeah, it's interesting. And then you have to like classify whether the page is, you have to decide which pages should be updated and what frequency. Oh yeah.Jeff Dean [00:29:30]: There's a whole like, uh, system behind the scenes that's trying to decide update rates and importance of the pages. So even if the update rate seems low, you might still want to recrawl important pages quite often because, uh, the likelihood they change might be low, but the value of having updated is high.Shawn Wang [00:29:50]: Yeah, yeah, yeah, yeah. Uh, well, you know, yeah. This, uh, you know, mention of latency and, and saving things to this reminds me of one of your classics, which I have to bring up, which is latency numbers. Every programmer should know, uh, was there a, was it just a, just a general story behind that? Did you like just write it down?Jeff Dean [00:30:06]: I mean, this has like sort of eight or 10 different kinds of metrics that are like, how long does a cache mistake? How long does branch mispredict take? How long does a reference domain memory take? How long does it take to send, you know, a packet from the U S to the Netherlands or something? Um,Shawn Wang [00:30:21]: why Netherlands, by the way, or is it, is that because of Chrome?Jeff Dean [00:30:25]: Uh, we had a data center in the Netherlands, um, so, I mean, I think this gets to the point of being able to do the back of the envelope calculations. So these are sort of the raw ingredients of those, and you can use them to say, okay, well, if I need to design a system to do image search and thumb nailing or something of the result page, you know, how, what I do that I could pre-compute the image thumbnails. I could like. Try to thumbnail them on the fly from the larger images. What would that do? How much dis bandwidth than I need? How many des seeks would I do? Um, and you can sort of actually do thought experiments in, you know, 30 seconds or a minute with the sort of, uh, basic, uh, basic numbers at your fingertips. Uh, and then as you sort of build software using higher level libraries, you kind of want to develop the same intuitions for how long does it take to, you know, look up something in this particular kind of.Shawn Wang [00:31:21]: I'll see you next time.Shawn Wang [00:31:51]: Which is a simple byte conversion. That's nothing interesting. I wonder if you have any, if you were to update your...Jeff Dean [00:31:58]: I mean, I think it's really good to think about calculations you're doing in a model, either for training or inference.Jeff Dean [00:32:09]: Often a good way to view that is how much state will you need to bring in from memory, either like on-chip SRAM or HBM from the accelerator. Attached memory or DRAM or over the network. And then how expensive is that data motion relative to the cost of, say, an actual multiply in the matrix multiply unit? And that cost is actually really, really low, right? Because it's order, depending on your precision, I think it's like sub one picodule.Shawn Wang [00:32:50]: Oh, okay. You measure it by energy. Yeah. Yeah.Jeff Dean [00:32:52]: Yeah. I mean, it's all going to be about energy and how do you make the most energy efficient system. And then moving data from the SRAM on the other side of the chip, not even off the off chip, but on the other side of the same chip can be, you know, a thousand picodules. Oh, yeah. And so all of a sudden, this is why your accelerators require batching. Because if you move, like, say, the parameter of a model from SRAM on the, on the chip into the multiplier unit, that's going to cost you a thousand picodules. So you better make use of that, that thing that you moved many, many times with. So that's where the batch dimension comes in. Because all of a sudden, you know, if you have a batch of 256 or something, that's not so bad. But if you have a batch of one, that's really not good.Shawn Wang [00:33:40]: Yeah. Yeah. Right.Jeff Dean [00:33:41]: Because then you paid a thousand picodules in order to do your one picodule multiply.Shawn Wang [00:33:46]: I have never heard an energy-based analysis of batching.Jeff Dean [00:33:50]: Yeah. I mean, that's why people batch. Yeah. Ideally, you'd like to use batch size one because the latency would be great.Shawn Wang [00:33:56]: The best latency.Jeff Dean [00:33:56]: But the energy cost and the compute cost inefficiency that you get is quite large. So, yeah.Shawn Wang [00:34:04]: Is there a similar trick like, like, like you did with, you know, putting everything in memory? Like, you know, I think obviously NVIDIA has caused a lot of waves with betting very hard on SRAM with Grok. I wonder if, like, that's something that you already saw with, with the TPUs, right? Like that, that you had to. Uh, to serve at your scale, uh, you probably sort of saw that coming. Like what, what, what hardware, uh, innovations or insights were formed because of what you're seeing there?Jeff Dean [00:34:33]: Yeah. I mean, I think, you know, TPUs have this nice, uh, sort of regular structure of 2D or 3D meshes with a bunch of chips connected. Yeah. And each one of those has HBM attached. Um, I think for serving some kinds of models, uh, you know, you, you pay a lot higher cost. Uh, and time latency, um, bringing things in from HBM than you do bringing them in from, uh, SRAM on the chip. So if you have a small enough model, you can actually do model parallelism, spread it out over lots of chips and you actually get quite good throughput improvements and latency improvements from doing that. And so you're now sort of striping your smallish scale model over say 16 or 64 chips. Uh, but as if you do that and it all fits in. In SRAM, uh, that can be a big win. So yeah, that's not a surprise, but it is a good technique.Alessio Fanelli [00:35:27]: Yeah. What about the TPU design? Like how much do you decide where the improvements have to go? So like, this is like a good example of like, is there a way to bring the thousand picojoules down to 50? Like, is it worth designing a new chip to do that? The extreme is like when people say, oh, you should burn the model on the ASIC and that's kind of like the most extreme thing. How much of it? Is it worth doing an hardware when things change so quickly? Like what was the internal discussion? Yeah.Jeff Dean [00:35:57]: I mean, we, we have a lot of interaction between say the TPU chip design architecture team and the sort of higher level modeling, uh, experts, because you really want to take advantage of being able to co-design what should future TPUs look like based on where we think the sort of ML research puck is going, uh, in some sense, because, uh, you know, as a hardware designer for ML and in particular, you're trying to design a chip starting today and that design might take two years before it even lands in a data center. And then it has to sort of be a reasonable lifetime of the chip to take you three, four or five years. So you're trying to predict two to six years out where, what ML computations will people want to run two to six years out in a very fast changing field. And so having people with interest. Interesting ML research ideas of things we think will start to work in that timeframe or will be more important in that timeframe, uh, really enables us to then get, you know, interesting hardware features put into, you know, TPU N plus two, where TPU N is what we have today.Shawn Wang [00:37:10]: Oh, the cycle time is plus two.Jeff Dean [00:37:12]: Roughly. Wow. Because, uh, I mean, sometimes you can squeeze some changes into N plus one, but, you know, bigger changes are going to require the chip. Yeah. Design be earlier in its lifetime design process. Um, so whenever we can do that, it's generally good. And sometimes you can put in speculative features that maybe won't cost you much chip area, but if it works out, it would make something, you know, 10 times as fast. And if it doesn't work out, well, you burned a little bit of tiny amount of your chip area on that thing, but it's not that big a deal. Uh, sometimes it's a very big change and we want to be pretty sure this is going to work out. So we'll do like lots of carefulness. Uh, ML experimentation to show us, uh, this is actually the, the way we want to go. Yeah.Alessio Fanelli [00:37:58]: Is there a reverse of like, we already committed to this chip design so we can not take the model architecture that way because it doesn't quite fit?Jeff Dean [00:38:06]: Yeah. I mean, you, you definitely have things where you're going to adapt what the model architecture looks like so that they're efficient on the chips that you're going to have for both training and inference of that, of that, uh, generation of model. So I think it kind of goes both ways. Um, you know, sometimes you can take advantage of, you know, lower precision things that are coming in a future generation. So you can, might train it at that lower precision, even if the current generation doesn't quite do that. Mm.Shawn Wang [00:38:40]: Yeah. How low can we go in precision?Jeff Dean [00:38:43]: Because people are saying like ternary is like, uh, yeah, I mean, I'm a big fan of very low precision because I think that gets, that saves you a tremendous amount of time. Right. Because it's picojoules per bit that you're transferring and reducing the number of bits is a really good way to, to reduce that. Um, you know, I think people have gotten a lot of luck, uh, mileage out of having very low bit precision things, but then having scaling factors that apply to a whole bunch of, uh, those, those weights. Scaling. How does it, how does it, okay.Shawn Wang [00:39:15]: Interesting. You, so low, low precision, but scaled up weights. Yeah. Huh. Yeah. Never considered that. Yeah. Interesting. Uh, w w while we're on this topic, you know, I think there's a lot of, um, uh, this, the concept of precision at all is weird when we're sampling, you know, uh, we just, at the end of this, we're going to have all these like chips that I'll do like very good math. And then we're just going to throw a random number generator at the start. So, I mean, there's a movement towards, uh, energy based, uh, models and processors. I'm just curious if you've, obviously you've thought about it, but like, what's your commentary?Jeff Dean [00:39:50]: Yeah. I mean, I think. There's a bunch of interesting trends though. Energy based models is one, you know, diffusion based models, which don't sort of sequentially decode tokens is another, um, you know, speculative decoding is a way that you can get sort of an equivalent, very small.Shawn Wang [00:40:06]: Draft.Jeff Dean [00:40:07]: Batch factor, uh, for like you predict eight tokens out and that enables you to sort of increase the effective batch size of what you're doing by a factor of eight, even, and then you maybe accept five or six of those tokens. So you get. A five, a five X improvement in the amortization of moving weights, uh, into the multipliers to do the prediction for the, the tokens. So these are all really good techniques and I think it's really good to look at them from the lens of, uh, energy, real energy, not energy based models, um, and, and also latency and throughput, right? If you look at things from that lens, that sort of guides you to. Two solutions that are gonna be, uh, you know, better from, uh, you know, being able to serve larger models or, you know, equivalent size models more cheaply and with lower latency.Shawn Wang [00:41:03]: Yeah. Well, I think, I think I, um, it's appealing intellectually, uh, haven't seen it like really hit the mainstream, but, um, I do think that, uh, there's some poetry in the sense that, uh, you know, we don't have to do, uh, a lot of shenanigans if like we fundamentally. Design it into the hardware. Yeah, yeah.Jeff Dean [00:41:23]: I mean, I think there's still a, there's also sort of the more exotic things like analog based, uh, uh, computing substrates as opposed to digital ones. Uh, I'm, you know, I think those are super interesting cause they can be potentially low power. Uh, but I think you often end up wanting to interface that with digital systems and you end up losing a lot of the power advantages in the digital to analog and analog to digital conversions. You end up doing, uh, at the sort of boundaries. And periphery of that system. Um, I still think there's a tremendous distance we can go from where we are today in terms of energy efficiency with sort of, uh, much better and specialized hardware for the models we care about.Shawn Wang [00:42:05]: Yeah.Alessio Fanelli [00:42:06]: Um, any other interesting research ideas that you've seen, or like maybe things that you cannot pursue a Google that you would be interested in seeing researchers take a step at, I guess you have a lot of researchers. Yeah, I guess you have enough, but our, our research.Jeff Dean [00:42:21]: Our research portfolio is pretty broad. I would say, um, I mean, I think, uh, in terms of research directions, there's a whole bunch of, uh, you know, open problems and how do you make these models reliable and able to do much longer, kind of, uh, more complex tasks that have lots of subtasks. How do you orchestrate, you know, maybe one model that's using other models as tools in order to sort of build, uh, things that can accomplish, uh, you know, much more. Yeah. Significant pieces of work, uh, collectively, then you would ask a single model to do. Um, so that's super interesting. How do you get more verifiable, uh, you know, how do you get RL to work for non-verifiable domains? I think it's a pretty interesting open problem because I think that would broaden out the capabilities of the models, the improvements that you're seeing in both math and coding. Uh, if we could apply those to other less verifiable domains, because we've come up with RL techniques that actually enable us to do that. Uh, effectively, that would, that would really make the models improve quite a lot. I think.Alessio Fanelli [00:43:26]: I'm curious, like when we had Noam Brown on the podcast, he said, um, they already proved you can do it with deep research. Um, you kind of have it with AI mode in a way it's not verifiable. I'm curious if there's any thread that you think is interesting there. Like what is it? Both are like information retrieval of JSON. So I wonder if it's like the retrieval is like the verifiable part. That you can score or what are like, yeah, yeah. How, how would you model that, that problem?Jeff Dean [00:43:55]: Yeah. I mean, I think there are ways of having other models that can evaluate the results of what a first model did, maybe even retrieving. Can you have another model that says, is this things, are these things you retrieved relevant? Or can you rate these 2000 things you retrieved to assess which ones are the 50 most relevant or something? Um, I think those kinds of techniques are actually quite effective. Sometimes I can even be the same model, just prompted differently to be a, you know, a critic as opposed to a, uh, actual retrieval system. Yeah.Shawn Wang [00:44:28]: Um, I do think like there, there is that, that weird cliff where like, it feels like we've done the easy stuff and then now it's, but it always feels like that every year. It's like, oh, like we know, we know, and the next part is super hard and nobody's figured it out. And, uh, exactly with this RLVR thing where like everyone's talking about, well, okay, how do we. the next stage of the non-verifiable stuff. And everyone's like, I don't know, you know, Ellen judge.Jeff Dean [00:44:56]: I mean, I feel like the nice thing about this field is there's lots and lots of smart people thinking about creative solutions to some of the problems that we all see. Uh, because I think everyone sort of sees that the models, you know, are great at some things and they fall down around the edges of those things and, and are not as capable as we'd like in those areas. And then coming up with good techniques and trying those. And seeing which ones actually make a difference is sort of what the whole research aspect of this field is, is pushing forward. And I think that's why it's super interesting. You know, if you think about two years ago, we were struggling with GSM, eight K problems, right? Like, you know, Fred has two rabbits. He gets three more rabbits. How many rabbits does he have? That's a pretty far cry from the kinds of mathematics that the models can, and now you're doing IMO and Erdos problems in pure language. Yeah. Yeah. Pure language. So that is a really, really amazing jump in capabilities in, you know, in a year and a half or something. And I think, um, for other areas, it'd be great if we could make that kind of leap. Uh, and you know, we don't exactly see how to do it for some, some areas, but we do see it for some other areas and we're going to work hard on making that better. Yeah.Shawn Wang [00:46:13]: Yeah.Alessio Fanelli [00:46:14]: Like YouTube thumbnail generation. That would be very helpful. We need that. That would be AGI. We need that.Shawn Wang [00:46:20]: That would be. As far as content creators go.Jeff Dean [00:46:22]: I guess I'm not a YouTube creator, so I don't care that much about that problem, but I guess, uh, many people do.Shawn Wang [00:46:27]: It does. Yeah. It doesn't, it doesn't matter. People do judge books by their covers as it turns out. Um, uh, just to draw a bit on the IMO goal. Um, I'm still not over the fact that a year ago we had alpha proof and alpha geometry and all those things. And then this year we were like, screw that we'll just chuck it into Gemini. Yeah. What's your reflection? Like, I think this, this question about. Like the merger of like symbolic systems and like, and, and LMS, uh, was a very much core belief. And then somewhere along the line, people would just said, Nope, we'll just all do it in the LLM.Jeff Dean [00:47:02]: Yeah. I mean, I think it makes a lot of sense to me because, you know, humans manipulate symbols, but we probably don't have like a symbolic representation in our heads. Right. We have some distributed representation that is neural net, like in some way of lots of different neurons. And activation patterns firing when we see certain things and that enables us to reason and plan and, you know, do chains of thought and, you know, roll them back now that, that approach for solving the problem doesn't seem like it's going to work. I'm going to try this one. And, you know, in a lot of ways we're emulating what we intuitively think, uh, is happening inside real brains in neural net based models. So it never made sense to me to have like completely separate. Uh, discrete, uh, symbolic things, and then a completely different way of, of, uh, you know, thinking about those things.Shawn Wang [00:47:59]: Interesting. Yeah. Uh, I mean, it's maybe seems obvious to you, but it wasn't obvious to me a year ago. Yeah.Jeff Dean [00:48:06]: I mean, I do think like that IMO with, you know, translating to lean and using lean and then the next year and also a specialized geometry model. And then this year switching to a single unified model. That is roughly the production model with a little bit more inference budget, uh, is actually, you know, quite good because it shows you that the capabilities of that general model have improved dramatically and, and now you don't need the specialized model. This is actually sort of very similar to the 2013 to 16 era of machine learning, right? Like it used to be, people would train separate models for lots of different, each different problem, right? I have, I want to recognize street signs and something. So I train a street sign. Recognition recognition model, or I want to, you know, decode speech recognition. I have a speech model, right? I think now the era of unified models that do everything is really upon us. And the question is how well do those models generalize to new things they've never been asked to do and they're getting better and better.Shawn Wang [00:49:10]: And you don't need domain experts. Like one of my, uh, so I interviewed ETA who was on, who was on that team. Uh, and he was like, yeah, I, I don't know how they work. I don't know where the IMO competition was held. I don't know the rules of it. I just trained the models, the training models. Yeah. Yeah. And it's kind of interesting that like people with these, this like universal skill set of just like machine learning, you just give them data and give them enough compute and they can kind of tackle any task, which is the bitter lesson, I guess. I don't know. Yeah.Jeff Dean [00:49:39]: I mean, I think, uh, general models, uh, will win out over specialized ones in most cases.Shawn Wang [00:49:45]: Uh, so I want to push there a bit. I think there's one hole here, which is like, uh. There's this concept of like, uh, maybe capacity of a model, like abstractly a model can only contain the number of bits that it has. And, uh, and so it, you know, God knows like Gemini pro is like one to 10 trillion parameters. We don't know, but, uh, the Gemma models, for example, right? Like a lot of people want like the open source local models that are like that, that, that, and, and, uh, they have some knowledge, which is not necessary, right? Like they can't know everything like, like you have the. The luxury of you have the big model and big model should be able to capable of everything. But like when, when you're distilling and you're going down to the small models, you know, you're actually memorizing things that are not useful. Yeah. And so like, how do we, I guess, do we want to extract that? Can we, can we divorce knowledge from reasoning, you know?Jeff Dean [00:50:38]: Yeah. I mean, I think you do want the model to be most effective at reasoning if it can retrieve things, right? Because having the model devote precious parameter space. To remembering obscure facts that could be looked up is actually not the best use of that parameter space, right? Like you might prefer something that is more generally useful in more settings than this obscure fact that it has. Um, so I think that's always attention at the same time. You also don't want your model to be kind of completely detached from, you know, knowing stuff about the world, right? Like it's probably useful to know how long the golden gate be. Bridges just as a general sense of like how long are bridges, right? And, uh, it should have that kind of knowledge. It maybe doesn't need to know how long some teeny little bridge in some other more obscure part of the world is, but, uh, it does help it to have a fair bit of world knowledge and the bigger your model is, the more you can have. Uh, but I do think combining retrieval with sort of reasoning and making the model really good at doing multiple stages of retrieval. Yeah.Shawn Wang [00:51:49]: And reasoning through the intermediate retrieval results is going to be a, a pretty effective way of making the model seem much more capable, because if you think about, say, a personal Gemini, yeah, right?Jeff Dean [00:52:01]: Like we're not going to train Gemini on my email. Probably we'd rather have a single model that, uh, we can then use and use being able to retrieve from my email as a tool and have the model reason about it and retrieve from my photos or whatever, uh, and then make use of that and have multiple. Um, you know, uh, stages of interaction. that makes sense.Alessio Fanelli [00:52:24]: Do you think the vertical models are like, uh, interesting pursuit? Like when people are like, oh, we're building the best healthcare LLM, we're building the best law LLM, are those kind of like short-term stopgaps or?Jeff Dean [00:52:37]: No, I mean, I think, I think vertical models are interesting. Like you want them to start from a pretty good base model, but then you can sort of, uh, sort of viewing them, view them as enriching the data. Data distribution for that particular vertical domain for healthcare, say, um, we're probably not going to train or for say robotics. We're probably not going to train Gemini on all possible robotics data. We, you could train it on because we want it to have a balanced set of capabilities. Um, so we'll expose it to some robotics data, but if you're trying to build a really, really good robotics model, you're going to want to start with that and then train it on more robotics data. And then maybe that would. It's multilingual translation capability, but improve its robotics capabilities. And we're always making these kind of, uh, you know, trade-offs in the data mix that we train the base Gemini models on. You know, we'd love to include data from 200 more languages and as much data as we have for those languages, but that's going to displace some other capabilities of the model. It won't be as good at, um, you know, Pearl programming, you know, it'll still be good at Python programming. Cause we'll include it. Enough. Of that, but there's other long tail computer languages or coding capabilities that it may suffer on or multi, uh, multimodal reasoning capabilities may suffer. Cause we didn't get to expose it to as much data there, but it's really good at multilingual things. So I, I think some combination of specialized models, maybe more modular models. So it'd be nice to have the capability to have those 200 languages, plus this awesome robotics model, plus this awesome healthcare, uh, module that all can be knitted together to work in concert and called upon in different circumstances. Right? Like if I have a health related thing, then it should enable using this health module in conjunction with the main base model to be even better at those kinds of things. Yeah.Shawn Wang [00:54:36]: Installable knowledge. Yeah.Jeff Dean [00:54:37]: Right.Shawn Wang [00:54:38]: Just download as a, as a package.Jeff Dean [00:54:39]: And some of that installable stuff can come from retrieval, but some of it probably should come from preloaded training on, you know, uh, a hundred billion tokens or a trillion tokens of health data. Yeah.Shawn Wang [00:54:51]: And for listeners, I think, uh, I will highlight the Gemma three end paper where they, there was a little bit of that, I think. Yeah.Alessio Fanelli [00:54:56]: Yeah. I guess the question is like, how many billions of tokens do you need to outpace the frontier model improvements? You know, it's like, if I have to make this model better healthcare and the main. Gemini model is still improving. Do I need 50 billion tokens? Can I do it with a hundred, if I need a trillion healthcare tokens, it's like, they're probably not out there that you don't have, you know, I think that's really like the.Jeff Dean [00:55:21]: Well, I mean, I think healthcare is a particularly challenging domain, so there's a lot of healthcare data that, you know, we don't have access to appropriately, but there's a lot of, you know, uh, healthcare organizations that want to train models on their own data. That is not public healthcare data, uh, not public health. But public healthcare data. Um, so I think there are opportunities there to say, partner with a large healthcare organization and train models for their use that are going to be, you know, more bespoke, but probably, uh, might be better than a general model trained on say, public data. Yeah.Shawn Wang [00:55:58]: Yeah. I, I believe, uh, by the way, also this is like somewhat related to the language conversation. Uh, I think one of your, your favorite examples was you can put a low resource language in the context and it just learns. Yeah.Jeff Dean [00:56:09]: Oh, yeah, I think the example we used was Calamon, which is truly low resource because it's only spoken by, I think 120 people in the world and there's no written text.Shawn Wang [00:56:20]: So, yeah. So you can just do it that way. Just put it in the context. Yeah. Yeah. But I think your whole data set in the context, right.Jeff Dean [00:56:27]: If you, if you take a language like, uh, you know, Somali or something, there is a fair bit of Somali text in the world that, uh, or Ethiopian Amharic or something, um, you know, we probably. Yeah. Are not putting all the data from those languages into the Gemini based training. We put some of it, but if you put more of it, you'll improve the capabilities of those models.Shawn Wang [00:56:49]: Yeah.Jeff Dean [00:56:49]:

Sousa Tavares analisa a resposta de Montenegro às tempestades: “deve ser o mais parecido que tivemos com uma guerra desde as invasões francesas”, para considerar que começou por faltar liderança e que foi o PR quem “puxou a carroça”. Sobre os efeitos, teme consequências económicas graves, propõe que a reconstrução seja feita a ter em conta os erros do passado e critica o momento escolhido pela MAI para deixar as funções. Sobre as presidenciais, diz que os eleitores “fizeram de um dia cinzento, um dia claro”. Fala ainda do papel de Seguro e deixa uma ideia em jeito de provocação sobre o voto dos emigrantes. Por fim, elogios para uma “excelente notícia” que chega da AR.See omnystudio.com/listener for privacy information.

Is Musicboard shutting down? Company says no, but users are worried. Also, Ouster is paying $35 million along with 1.8 million shares. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Episode 210: In this episode of Accelerate, host Nicola Graham is joined by Emma Meehan — Founder, CEO, and CTO of KinetikIQ. Emma is building technology that sits at the intersection of biomechanics, machine learning, and real-world performance. KinetikIQ turns any smartphone into a full-body 3D biomechanics system using LiDAR and AI — no wearables required — making advanced movement analysis far more accessible across sport and health. With a background in computer science and software engineering, alongside experience as a competitive weightlifter, Emma brings both technical depth and practitioner perspective to product development. Her work has already been recognised across sport, technology, and business — including wins at the KPMG Global Tech Innovator Ireland and the Barca Innovation Challenge, Best New Sports Business of the Year at the Irish Sport Awards, recognition from SportsTechX as a European startup to watch, and features in the Sunday Business Post and Irish Independent 30 Under 30 lists. Together, Nicola and Emma explore what it really takes to build a company as a technical founder, how the Irish startup ecosystem can support early-stage growth, and the realities of securing venture capital in sport and healthtech — alongside the lived experience of building as a female founder in a still-emerging industry. Topics discussed: Building a company as a technical founder The role of the Irish startup ecosystem in early growth Venture capital funding in sport and healthtech The realities of being a female founder in sports technology Where you can find Emma: LinkedIn Instagram KineticIQ - Sponsors Gameplan is a rehab Project Management & Data Analytics Platform that improves operational & communication efficiency during rehab. Gameplan provides a centralised tool for MDT's to work collaboratively inside a data rich environment VALD Performance, makers of the ForceDecks, ForceFrame, HumanTrak, Dynamo, SmartSpeed, NordBoard. VALD Performance systems are built with the high-performance practitioner in mind, translating traditionally lab-based technologies into engaging, quick, easy-to-use tools for daily testing, monitoring and training Hytro: The world's leading Blood Flow Restriction (BFR) wearable, designed to accelerate recovery and maximise athletic potential using Hytro BFR for Professional Sport.  -  Where to Find Us Keep up to date with everything that is going on with the podcast by following Inform Performance on: Instagram Twitter Our Website - Our Team Andy McDonald Ben Ashworth Steve Barrett  Pete McKnight

After Ouster (OUST) announced its acquisition of StereoLabs, the stock jumped 10% on Monday's session. The company's CEO and co-founder, Angus Pacala, explains how the acquisition allows Ouster to build a "unified" platform combining AI compute, cameras, and LiDAR in its autonomous tech. StereoLabs CEO Cecile Schmollgruber talks about how her company built the camera technology by studying human vision. ======== Schwab Network ========Empowering every investor and trader, every market day.Options involve risks and are not suitable for all investors. Before trading, read the Options Disclosure Document. http://bit.ly/2v9tH6DSubscribe to the Market Minute newsletter - https://schwabnetwork.com/subscribeDownload the iOS app - https://apps.apple.com/us/app/schwab-network/id1460719185Download the Amazon Fire Tv App - https://www.amazon.com/TD-Ameritrade-Network/dp/B07KRD76C7Watch on Sling - https://watch.sling.com/1/asset/191928615bd8d47686f94682aefaa007/watchWatch on Vizio - https://www.vizio.com/en/watchfreeplus-exploreWatch on DistroTV - https://www.distro.tv/live/schwab-network/Follow us on X – https://twitter.com/schwabnetworkFollow us on Facebook – https://www.facebook.com/schwabnetworkFollow us on LinkedIn - https://www.linkedin.com/company/schwab-network/About Schwab Network - https://schwabnetwork.com/about

La tecnología LiDAR revela caminos ocultos de la Ruta del Ámbar, una antigua red comercial que unía el Báltico con el Mediterráneo, y transforma nuestra visión de la Historia.Un hallazgo revolucionario en Newgrange descubre la existencia de poderosas líneas dinásticas y revela prácticas incestuosas de la élite dirigente que conectan a los antiguos soberanos de Irlanda.

This week on Autonomy Markets, Grayson Brulte and Walter Piecyk are joined by their first-ever guest, David Moss, to discuss his 12,961-mile zero-intervention drive across the country on Tesla FSD, the reality of the Unsupervised Robotaxi rollout in Austin, and the commercial viability of LiDAR sensors in consumer vehicles.The conversation heats up as Walt questions David, a LiDAR LiDAR Salesman on whether the massive data processing requirements of LiDAR could introduce latency, potentially citing a recent Waymo incident involving a child as a case study. David argues that while LiDAR offers theoretical range advantages, the compute wall and cost constraints make it a one-trick pony compared to the scalability of a vision-only stack.While the group debates sensor suites, David shares his on-the-ground experience in Austin, revealing it took 58 attempts to finally secure a ride in a Unsupervised Tesla Robotaxi, and confirmed the fleet is being retrofitted with new cleaning jets for the camera sensors to handle weather occlusion.Looking at the broader robotaxi market, the trio analyzes their Zoox experiences at CES, with David noting the vehicle's braking was significantly harsher than Waymo or Tesla FSD, while Walt highlights the motion sickness challenges inherent in the vehicle's carriage-style seating configuration.In Prediction Corner, the group debates the timeline for Tesla removing the safety driver on highways, with David offering a bullish forecast for Memorial Day, while Walt and Grayson take a more conservative stance, predicting a rollout closer to late 2026.Episode Chapters0:00 Coast to Coast Fully Autonomous in a Tesla Model 310:49 The Next Record12:16 FSD Unsupervised in Austin16:16 Waymo Experience on Uber in Austin17:17 Robotaxi Safety Attendants19:44 Unsupervised Robotaxi Service Area21:43 Sensor Cleaning26:05 Robotaxi, No Highways in Austin, Yet32:11 Zoox Las Vegas Experiences37:13 LiDAR48:07 Why AutonomyRecorded on Tuesday, February 3, 2026 --------About The Road to AutonomyThe Road to Autonomy provides market intelligence and strategic advisory services to institutional investors and companies, delivering insights needed to stay ahead of emerging trends in the autonomy economy™. To learn more, say hello (at) roadtoautonomy.com.Sign up for This Week in The Autonomy Economy newsletter: https://www.roadtoautonomy.com/ae/See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

This week the Geoholics crew sits down with Rob Cammack — entrepreneur, builder, and CEO/Founder of SmartDrone — a guy who's basically been innovating since most of us were trading baseball cards. From launching businesses as a kid to founding and selling multiple companies, Rob brings that rare mix of grit, vision, and “there's gotta be a better way” thinking that our industry desperately needs. We dive headfirst into the origin story behind Magellan, the first “No Drone Expert Required” LiDAR drone — and why it's flipping the script on how survey firms adopt technology. No six-figure science projects. No “drone guy risk.” Just practical robotics that turn everyday crews into productivity machines. But this episode isn't just about hardware — it's about mindset. Rob talks: -Building companies (and selling them) without losing your soul -Removing friction so surveyors can focus on surveying -Why LiDAR is the great unlock for mapping pros stuck in camera-only workflows -How robotics can upgrade people, not replace them -What “world-class talent” really looks like in a tech-driven survey company -And where the surveying industry is headed in the next 5–10 years It's equal parts leadership, innovation, and straight-up field practicality — the kind of conversation that makes you rethink how your team operates Monday morning. If you care about LiDAR, drones, scaling your survey operations, or building a future-proof business… this one's a must-listen. Because sometimes innovation isn't about flying higher…it's about making it simple enough that everyone can fly. Music by The Killers!

The gang returns for 2026 with CES chat, and predictions that will age like milk!  What's inside: Kayaker freezes in New England; Sean freezes in the original England; Garth refuses to stop being sunny Steven Scott drops in from Double Tap and immediately turns "a quick chat" into "a feature-length film" AirPods Pro 3 praise, ear-tip rage, and the haunting tale of the "dog-processed" AirPods The Lightsaber Cane: brilliant, ridiculous, heavier than normal, and only sometimes practical "Does this confuse the public?" Spoiler: the public is already confused Keychron keyboards, accessibility quirks, and a CES announcement of a concrete keyboard (because why not) HP's keyboard-with-a-computer concept:  Robot future: great in factories, questionable in living rooms, nightmare fuel for kids' toys AI glasses talk: open ecosystems, camera access, LiDAR dreams, and "Meta, please stop resetting" Braille label printer news: useful… if the app isn't a disaster 2026 predictions: Siri overhaul (again), Google wearables rising, AI hype deflating, and general tech chaos Overall, CES + accessibility + Star Wars + keyboards + mild existential dread. Exactly what you came for.

This week on Autonomy Markets, Grayson Brulte and Walter Piecyk discuss Waymo's LiDAR controversy following an incident in a Santa Monica school zone, the potential of a mega merger between Tesla, SpaceX, and xAI, and Waabi's $750 million capital raise to pivot into robotaxis.The conversation heats up as Walt and Grayson debate the efficacy of LiDAR versus camera-only approaches, questioning if sensor fusion latency contributed to the Waymo incident where a child ran out from behind a vehicle.While Waymo handles the incident in Santa Monica, Tesla is further accelerating their shift to an autonomy/robotics company by shutting down Model S and Model X lines for Optimus Gen 3 humanoids.Looking at the broader market, Grayson and Walt analyze Waabi's strategic expansion from trucking into robotaxis, with Walt drawing parallels to early industry pivots and Grayson questioning the viability of managing two distinct autonomy programs.On the Foreign Autonomy Desk, they highlight Waymo's recent launch party in London, noting the imported vehicles still feature American driving configurations, and discuss Pony.ai's partnership to deploy 3,000 robotaxis in mainland China.Episode Chapters0:00 Waymo Opens SFO Airport Access5:45 Waymo's Santa Monica Incident16:27 Tesla Earnings and New Robotaxi Markets22:04 David Moss' Austin Robotaxi Adventures24:12 Robotaxi's Enhanced Camera Cleaning System26:38 Inspector Uncovers, Walt Warns27:43 The Potential Great Elon Merger30:35 Waabi Raises $750m, Pivots to Robotaxis39:27 Does Uber Reboot ATG?42:55 Plus AI Analyst Day50:01 Foreign Autonomy Desk53:18 Next WeekRecorded on Friday, January 30, 2026--------About The Road to AutonomyThe Road to Autonomy provides market intelligence and strategic advisory services to institutional investors and companies, delivering insights needed to stay ahead of emerging trends in the autonomy economy™. To learn more, say hello (at) roadtoautonomy.com.Sign up for This Week in The Autonomy Economy newsletter: https://www.roadtoautonomy.com/ae/See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

In this episode, Dominic Rubino talks with Steven Moran (CEO of FlexiJet Digital Measuring) about how digital measuring can solve "people problems" by improving data accuracy, making delegation easier, and reducing rework. In this episode, we cover: Why measurement errors create costly go-backs How digital measuring reduces manual entry mistakes Training newer team members faster (without 20 years experience) Capturing jobsite info like "forensics" (photo + documentation) How better systems support succession planning and scaling What's coming next (LiDAR + future workflows)

In this episode, we sit down with Greg McCunn, President and CEO of Great Pacific Gold (TSXV: GPAC | OTCQX: GPGCF), to discuss the company's aggressive 2026 exploration strategy at the Wild Dog Project, in Papua New Guinea. Following the release of high-grade results from the Sinivit area, Greg outlines how the company is transitioning from localized discovery to district-scale expansion. Key Discussion Points Expanding the Northern Sulfide Zone - Greg details the significance of Hole 15, which returned 13.5 meters at 8.1 g/t AuEq, proving that mineralization remains open and potentially increases in grade at depth. District-Scale Potential - Insights into the Wild Dog structural corridor, where current drilling at Sinivit covers only 10% of the 15-kilometer strike length. The 2026 Drill Program - A breakdown of the 10,000 to 20,000-meter drill campaign designed to test high-priority targets including Kasie Ridge, Kavasuki and Mengmu. New Target Generation - How the company utilizes LIDAR and Mobile MT geophysics to move 25 distinct targets through the exploration pipeline toward drill-ready status.   If you have any follow up questions for Greg please me at Fleck@kereport.com.  Click here to visit the Great Pacific Gold website - https://gpacgold.com/   ------------------------ For more market commentary & interview summaries, subscribe to our Substacks:  The KE Report: https://kereport.substack.com/  Shad's resource market commentary: https://excelsiorprosperity.substack.com/   Investment disclaimer: This content is for informational and educational purposes only and does not constitute investment advice, an offer, or a solicitation to buy or sell any security or investment product. Investing in equities, commodities, really everything involves risk, including the possible loss of principal. Do your own research and consult a licensed financial advisor before making any investment decisions. Guests and hosts may own shares in companies mentioned.

Uber is not developing its own robotaxis again; instead it plans to collect and offer data. It's a bet that more volume will help autonomous vehicle partners solve the weirdest edge cases. Also, a new leading bidder has appeared in the Luminar bankruptcy case: Redmond, Washington-based MicroVision, which beat out Quantum Computing Inc.'s bid by $5 million. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Today's guests are Daniel Lidar, Holder of the Viterbi Professorship of Engineering at USC, Director of the USC Center for Quantum Information Science & Technology, co-founder and CSO of Quantum Elements, Inc., and Izhar Medalsy, Co-founder and CEO of Quantum Elements. Quantum Elements develops tools to reduce noise in quantum computers for scalable performance. Daniel and Izhar join Emerj Editorial Director Matthew DeMello to explore why quantum computing is entering enterprise strategy now, explaining qubits, quantum simulation, and practical applications beyond hype. Daniel and Izhar also share practical takeaways like quantum simulation for accurate materials and drug design, optimization for financial portfolios, post-quantum encryption for data centers, and assessing business impact by identifying problems quantum solves today. Want to share your AI adoption story with executive peers? Click emerj.com/expert2 for more information and to be a potential future guest on the 'AI in Business' podcast!

In this episode of The Geoholics Podcast, the crew dives deep—both literally and figuratively—into the world of archaeology, GIS, and cultural resource management with special guest Jenna Kent, Archaeologist at Jacobs Engineering Group. From growing up across Texas, Mississippi, Utah, and Hawaii as part of a military family, to excavating 7th-century monasteries and 12th-century abbeys in Ireland, Jenna's journey has been anything but ordinary. That geographic diversity helped shape her appreciation for landscapes, cultures, and the human stories hidden beneath them. The conversation explores what archaeology really looks like beyond the movies—balancing rugged fieldwork with complex office analysis—and why cultural resource compliance is far more technical, analytical, and geospatially driven than most people realize. Listeners get an inside look at: >Prehistoric ceramic replication and how recreating ancient pottery reveals insights no textbook ever could >Surveying 15 miles of wilderness at Bandelier National Monument, one of Jenna's career-defining projects >How archaeologists decode fragmented evidence like a massive puzzle with missing pieces >The growing role of GIS in archaeology, including site density modeling, probability mapping, and interactive story maps >Where surveyors, mappers, LiDAR professionals, and archaeologists can collaborate more effectively >The powerful human moments that remind us archaeology is ultimately about people—not artifacts Jenna closes the episode with thoughtful advice for young professionals looking to enter archaeology, cultural resources, or GIS—encouraging curiosity, patience, and a willingness to embrace both science and storytelling. This episode is a reminder that whether you're mapping terrain, scanning infrastructure, or excavating history—context matters, layers matter, and collaboration across disciplines makes us all better. Song of the Week: “New Orleans Is Sinking” by The Tragically Hip  

On today's highly observant episode of Quick Charge, Waymo founder John Krafcik takes aim at Tesla's Full Self Driving hardware limitations and Volvo Cars rolls out their most important new product of the 2020s: the all-new EX60 electric SUV! The Waymo founder says it's Tesla's antiquated camera tech, not necessarily its FSD software, that's keeping the company from offering truly autonomous robotaxis – and even says they'd fail a DMV eye exam! We've also got a look at the all-new, ultra fast charging Volvo EX60 and Peter Johnson looks into the crystal ball to peer into the future of Hyundai's upscale Genesis brand. Source Links Waymo founder John Krafcik: Tesla's Full Self-Driving has ‘bad case of myopia' Tesla patents ‘clever math trick' for HW3, but nothing points to delivering promised self-driving Tesla quietly cuts 1,700 jobs at Gigafactory Berlin despite denying it Volvo reveals EX60 SUV, its fastest charging EV yet – and an offroad surprise Volvo set to ditch LiDAR for 2026 – and Luminar is BIG mad Genesis outsold Infiniti in the US in 2025, now it's closing in on Lincoln and Acura Genesis emerges as a dark horse in the luxury EV space as even bigger plans unfold Genesis secretly designed this electric pickup and may bring it to life [Images] Prefer listening to your podcasts? Audio-only versions of Quick Charge are now available on Apple Podcasts, Spotify, TuneIn, and our RSS feed for Overcast and other podcast players. New episodes of Quick Charge are (allegedly) recorded several times per week, most weeks. We'll be posting bonus audio content from time to time as well, so be sure to follow and subscribe so you don't miss a minute of Electrek's high-voltage podcast series. Got news? Let us know!Drop us a line at tips@electrek.co. You can also rate us on Apple Podcasts and Spotify, or recommend us in Overcast to help more people discover the show. If you're considering going solar, it's always a good idea to get quotes from a few installers. To make sure you find a trusted, reliable solar installer near you that offers competitive pricing, check out EnergySage, a free service that makes it easy for you to go solar. It has hundreds of pre-vetted solar installers competing for your business, ensuring you get high-quality solutions and save 20-30% compared to going it alone. Plus, it's free to use, and you won't get sales calls until you select an installer and share your phone number with them.  Your personalized solar quotes are easy to compare online and you'll get access to unbiased Energy Advisors to help you every step of the way. Get started here.

Lidar com filhos adolescentes pode parecer uma batalha diária — silêncio, afastamento, explosões emocionais, conflitos que surgem do nada. Se você sente que perdeu o acesso ao seu filho ou que a comunicação dentro de casa ficou mais difícil, este episódio é para você.Aqui, a conversa vai além do comportamento. Falamos sobre o que realmente acontece por trás da adolescência: as transformações emocionais, neurológicas e internas que afetam não só os filhos, mas toda a família. Você vai entender por que essa fase ativa medos, inseguranças e gatilhos emocionais nos pais — e como atravessar esse período com mais consciência, presença e leveza.Ao longo do episódio, você vai perceber que educar um adolescente não é sobre controlar, corrigir ou impor autoridade, mas sobre aprender a sustentar vínculo, escuta e limites saudáveis sem perder a conexão. A reflexão é profunda, prática e aplicável ao dia a dia real, sem fórmulas prontas ou discursos rígidos.Se você busca melhorar o relacionamento com seus filhos, fortalecer o vínculo familiar, reduzir conflitos e atravessar a adolescência com mais clareza emocional, este episódio oferece um novo olhar — mais humano, mais consciente e possível.Dê o play com calma. Essa escuta pode transformar não só a relação com seu filho, mas também a forma como você se relaciona com você mesma.#adolescência #filhosepais #educaçãoemocional #consciênciafamiliar #relacionamentos #parentalidadeconsciente #inteligênciaemocional #família

WWJ auto analyst John McElroy reports one company has reached a breakthrough with artificial intelligence and radar signals.

Cotrim cheio de vidro, comprar pantufas, lidar com problemas - Fora da Lei #275 by Tiago Almeida

Can we really build a $10,000 humanoid robot on open-source AI?In this episode of TechFirst, John Koetsier talks with Chris Kudla, CEO of Mind Children, about a radically different approach to humanoid robots. Instead of six-figure industrial machines built for factories or war zones, Mind Children is building small, safe, friendly social robots designed for kids, classrooms, and elder care.Meet Cody (MC-1), their first humanoid prototype. Cody is built on open-source AI from SingularityNET, combined with modular hardware, low-torque actuators, and a wheeled base designed for safety, affordability, and mass production. And there's some other AI bits and pieces from all the big name companies that you'd recognize.Mind Children's goal is ambitious: a $10,000 humanoid robot that families, schools, and care facilities can actually afford.In this conversation we explore:• Why social robots may be the real gateway to embodied AI• How Cody is designed for children and elder care instead of factories• Why wheels beat bipedal legs for safety, cost, and stability• How open-source AI and modular software stacks enable faster innovation • The emotional and ethical challenges of building companion robots• And what it takes to bring a humanoid robot to market at scaleThis is not sci-fi. This is the early blueprint of a future where humanoid robots are personal, affordable, and open-source.00:00 – The $10,000 open-source humanoid question01:58 – Meet Cody, the MC-1 prototype04:10 – Why Cody is small, child-sized, and approachable06:55 – Designing humanoids for kids and elder care09:45 – Social robots vs industrial humanoids12:40 – Wheels instead of legs and why that matters16:05 – Low-torque actuators, safety, and toy-like design19:20 – Modular hands, arms, and future upgrades22:10 – Open-source AI and SingularityNET's role25:30 – On-robot vs cloud AI and why it matters28:40 – Vision, LiDAR, and simulated world models32:10 – Emotional awareness and social intelligence35:10 – The $10K target and mass-production strategy38:15 – The risks of attachment to robot companions40:00 – Final thoughts on Cody and the future of social robots

Studying the atmosphere from the sky can give us a clear picture of invisible hazards like turbulence or smoke.

Survivors underground in a major city need one thing more than anything else: Power. A war for power is in the making and in the mean time reliance on an A.I. is changing the way people do things in every way. Whether or not the AI is out of alignment depends on who you ask. Who is in control, people or the AI?Enviro-suits are wearable survival garments that cool the body, filter air, manage humidity, and protect against heat spikes and toxic environments.  Clear bell hoods are transparent helmet enclosures that integrate with enviro-suits to provide sealed breathing space and heads-up display support.  A.R. glasses are augmented reality eyewear that overlays digital information, navigation, and warnings onto the physical world.  A.R. night vision is a vision enhancement mode that allows people to see in darkness using sensor data rather than visible light.  Canal links are implanted or wearable communication and sensing devices that connect users to networks, A.R. systems, and AI services.  Wi-Fi sensing vision is a perception system that uses reflected radio waves to map environments and detect living beings through obstacles.  Interactive light clothing uses embedded LEDs and responsive fabrics to illuminate surroundings and signal presence.  LED headlamps and body lamps are personal lighting devices integrated into clothing or worn externally for navigation in dark environments.  Production centers are automated manufacturing hubs that 3D print new equipment, suits, devices, and consumer goods.  Two-dimensional material glass is ultra-strong non-silicate window material resistant to extreme winds and debris.  Robotic salvage bots are repurposed robots recovered from ruins and reassembled for labor, transport, or construction.  Hacked modular robots are custom-built machines assembled from mismatched salvaged parts and adapted through trial-and-error engineering.  Exoskeleton suits are powered wearable frames that enhance strength and can operate independently as robotic platforms.  Autono-carts are autonomous transport robots with legs instead of wheels, designed to move through flooded or uneven terrain.  Handy bots are humanoid utility robots used for general labor, security, and assistance tasks.  Follow carts are robotic cargo carriers that automatically trail their owner and transport goods.  Mag-soldering pliers are tools that use magnetic fields to hold and fuse electrical connections precisely.  Pseudo-superconductor wire is salvaged cabling capable of efficiently transmitting power and providing cooling effects.  Heat pump cooling units are compact devices reclaimed or rebuilt to regulate temperature inside suits and homes.  My-crete is a bio-engineered construction material grown or poured in place, used for walls, gates, and structural reinforcement.  Bail-block construction is a building method using compressed waste blocks insulated with my-crete for housing and workspaces.  Nuclear container reactors are compact nuclear power generators originally designed to power data centers and now repurposed for city energy.  Floating internet is a decentralized network infrastructure rebuilt after collapse to allow communication and data exchange.  Thrive is a powerful AI system that organizes labor, trade, safety, and social behavior through augmented reality guidance and incentives.  Assist AIs are personal digital helpers that provide navigation, alerts, communication, and decision support.  Jobs Navigator is a pre-collapse global AI system that Thrive is based on, designed to match people with work efficiently.  A.R. identity tagging flags individuals with trust, threat, or behavior markers visible only through augmented reality systems.  Spec-size key chips are implanted access control devices embedded in the body to unlock private living spaces.  Connected spectroscopy is a sensor system that analyzes materials, such as coins, to verify authenticity.  Lidar vision is a scanning technology used alongside A.R. to map spaces and detect objects with precision.  Encrypted digital currencies like Cashola are post-collapse monetary systems used to pay workers outside AI-controlled economies.  Coin authentication A.I. is a vision system capable of distinguishing real pre-collapse currency from printed fakes.  Crem makers are advanced food fabrication appliances that convert biological input into bread, meat, cheese, and other foods.  Yeast-meat technology is an earlier biofabrication method for producing protein foods through fermentation.  Battery reclamation systems collect, recharge, and redistribute discarded power cells as a core economic activity.  Automated security gates are layered physical defenses combining chain, composite materials, and robotic control.  Autono-shooters are automated defensive weapon systems used to deter or stop intruders.  Nano-wire traps are high-strength, nearly invisible defensive barriers designed to injure or stop attackers.  Live cam wearables are body-mounted cameras used by security teams to provide constant surveillance feeds.  Medical first-aid bots are autonomous robotic units that perform emergency surgery, drug delivery, and wound treatment.  Phage-based healing solutions use engineered viruses to accelerate tissue repair after injury.  Designed infections are engineered biological agents used to deliberately impair or control human function.  Parasite-based neuro-modulation uses modified single-cell organisms to alter perception, behavior, or cognition.  Methane collectors are bio-gas systems that harvest methane from anaerobic decomposition of plant waste.  Methane-powered generators are converted combustion engines that burn methane to produce electricity.  Magnetic cooling suits use current-driven magnetic fields in specialized cables to reduce body temperature.  A.R. navigation arrows are visual guides projected into a user's view to direct movement and tasks.  Micro-payment labor systems reward small acts of work or cooperation with fractional digital currency.  AI-managed retirement and income planning automatically allocates earnings and future resources for users.  Robotic inspection bots are machines tasked with scanning other robots for sabotage, tracking devices, or faults.  Lutin Two robots are commercially produced humanoid robots known for reliability and modern design.  Strider carts are multi-legged powered vehicles capable of navigating flooded tunnels, stairs, and debris-filled routes.Many of the characters in this project appear in future episodes.Using storytelling to place you in a time period, this series takes you, year by year, into the future. From 2040 to 2195. If you like emerging tech, eco-tech, futurism, perma-culture, apocalyptic survival scenarios, and disruptive science, sit back and enjoy short stories that showcase my research into how the future may play out. The companion site is https://in20xx.com These are works of fiction. Characters and groups are made-up and influenced by current events but not reporting facts about people or groups in the real world. This project is speculative fiction. These episodes are not about revealing what will be, but they are to excited the listener's wonder about what may come to pass.Copyright © Cy Porter 2025. All rights reserved.

In this episode of Tacos & Tech, Neal Bloom sits down with Eric Aguilar, co-founder and CEO of Omnitron Sensors, to unpack the deep tech powering the next wave of robotics, autonomous systems, and AI infrastructure. From Eric's early days working on defense sensors and his journey through Google and Tesla, to building one of the most powerful MEMS-based micro-machines on the market, this conversation explores why physical AI is finally having its moment - and what it takes to solve real-world reliability problems at scale.Eric breaks down why LiDAR has struggled to reach mass adoption, how Omnitron rethought the problem from first principles, and why the same core technology is now attracting attention from trillion-dollar hyperscalers looking to radically reduce data center power consumption. Along the way, they dive into biomimicry, energy efficiency, manufacturing constraints, and what it really means to build a “painkiller, not a vitamin.”Key Topics Covered* Eric's path from Navy research labs to Google, Tesla, and founding Omnitron Sensors* Why LiDAR reliability - not hype - has been the biggest blocker to autonomous systems* How MEMS-based silicon micro-machines replace failure-prone mechanical LiDAR components* The “war on LiDAR” and why cameras alone still fall short in autonomy* Omnitron's breakthrough in building large, high-force, highly reliable MEMS mirrors* Why physical AI and robotics are converging right now* How Omnitron's technology extends beyond automotive into AI data centers and optical switching* Saving massive amounts of energy by keeping data optical instead of electrical* The hidden challenges of scaling hardware, manufacturing, and global supply chains* Why MEMS has historically been underfunded - and why that's changing* Biomimicry as a tool for engineering breakthroughs* Advice for engineers and operators thinking about taking the leap into startups* Why the best startups solve urgent pain, not “nice-to-have” problems* Eric's very non-consensus taco pick (hint: it's a legendary San Diego burger)Links & Resources* Omnitron Sensors – https://www.omnitronsensors.comConnect with Eric* Eric Aguilar on LinkedIn This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit risingtidepartners.substack.com/subscribe

Discover the new AI Guide Dog app for iPhone, explore how LiDAR is finally being used for obstacle detection, and hear listener stories about blind dating, social experiences, and navigating accessibility in everyday life.Steven Scott and Shaun Preece dive into the AI Guide Dog iOS app, which turns your iPhone into a real-time obstacle detector using LiDAR or the front camera. Shaun shares how the app works with haptic feedback and distance alerts, comparing it to older mobility tools like the MiniGuide and Sunu Band. The conversation also explores practical uses, limitations, and whether apps like this can complement guide dogs or canes. Listener Charles responds with thoughtful insights about dating and relationships within the blind community, discussing personal experiences with sighted and blind partners, and how disability can influence—but not define—our connections. Sarah from the Netherlands shares tips for identifying and buying Nespresso Virtuo pods and highlights the brand's Braille labelling service. Finally, Jeff takes us on a nostalgic trip through early computing, punch cards, and green-screen terminals, reflecting on how far technology and accessibility have come.Enjoying Double Tap? Share your thoughts! Email feedback@doubletaponair.com or send a voice note via WhatsApp at +1 613-481-0144. Don't forget to like, subscribe, and leave a review to support accessible tech conversations. Find Double Tap online: YouTube, Double Tap Website---Follow on:YouTube: https://www.doubletaponair.com/youtubeX (formerly Twitter): https://www.doubletaponair.com/xInstagram: https://www.doubletaponair.com/instagramTikTok: https://www.doubletaponair.com/tiktokThreads: https://www.doubletaponair.com/threadsFacebook: https://www.doubletaponair.com/facebookLinkedIn: https://www.doubletaponair.com/linkedin Subscribe to the Podcast:Apple: https://www.doubletaponair.com/appleSpotify: https://www.doubletaponair.com/spotifyRSS: https://www.doubletaponair.com/podcastiHeadRadio: https://www.doubletaponair.com/iheart About Double TapHosted by the insightful duo, Steven Scott and Shaun Preece, Double Tap is a treasure trove of information for anyone who's blind or partially sighted and has a passion for tech. Steven and Shaun not only demystify tech, but they also regularly feature interviews and welcome guests from the community, fostering an interactive and engaging environment. Tune in every day of the week, and you'll discover how technology can seamlessly integrate into your life, enhancing daily tasks and experiences, even if your sight is limited. "Double Tap" is a registered trademark of Double Tap Productions Inc. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Kicking off 2026, Tu and Lei return from CES in Las Vegas with firsthand insights into how the global auto industry's center of gravity continues to shift toward China, AI, autonomy, and robotics.  This episode unpacks why CES is no longer about cars, but about who controls the software, silicon, sensors, and robots that will define the next decade of mobility. From Geely and Great Wall's growing U.S. ambitions, to Hyundai's robot-only keynote, to Ford's quiet but meaningful autonomy reset, the hosts connect dots that most headlines missed.Tu and Lei also break down the Geely “coming to the U.S.” scoop, Rivian-style AI days spreading to legacy OEMs, and why Western automakers are increasingly borrowing from China's playbook—from ADAS and silicon strategy to embodied AI and robotics.The episode closes with a deep dive into autonomy's three tracks (L2++, consumer L3/L4, and robotaxis), the growing importance of LiDAR scale, and why Donut Labs' solid-state battery and in-wheel motor reveal could become a true industry disruptor—if it scales.Fast, candid, and packed with on-the-ground context, this episode explains why CES 2026 marked a turning point—and why the race is no longer just EVs vs ICE, but ecosystems vs incumbents.___

SummaryFrom Westonbirt inspirations to field guides and plant-hunter epics, Lewis and Gemma pull 13 tree books and ask how reading changes woodland practice. Hear about ships with greenhouses, coppice cycles, charcoal burning, fungal networks, minimalist nursery design, mapping with old OS layers and LiDAR, plus a practitioner's starter stack for ID and ethnobotany.SponsorsTENTSILESave 10% on tree tents and hammocks with code ForestChildren10 at checkout. Ideal for leaders who want flexible base-camp shelter without ground impact.Chris HollandExplore Chris's 54-page Plant of the Week guide with songs, stories and QR videos. Use our affiliate link: https://chrisholland.myshopify.com/?ref=ForestSchoolPodcastKey takeawaysBooks are tools. Ideas on the page translate into better planning, richer invitations to play and clearer woodland decisions.History explains today's woods. War, trade and enclosure shaped plantations and access.When the landscape is the resource you can need fewer add-ons.Mycorrhizal science challenges the clean slate approach to plantations. Diversity can feed young trees.A balanced shelf helps practitioners. Mix narrative inspiration, technical ID, land-use history and local mapping.Chapters00:00 Audio or video and how to follow along02:10 Westonbirt, tree hunters and why one book leads to three more06:40 Plant collectors, ships with greenhouses and species introductions11:20 Remarkable trees and the Douglas fir story15:20 Finding the Mother Tree and what fungal networks show us20:10 Managing woods for play, coppice cycles and charcoal25:40 Enclosure, disafforestation and the Western Rising rabbit hole30:40 Rackham, old OS maps and first steps with LiDAR35:30 Practitioner stack for sessions and ethnobotany40:50 Photos or illustrations for ID, trends in tree writing, the squirrel book wishBooks and resources mentionedThomas Pakenham — The Tree Hunters; Meetings with Remarkable TreesJohn Evelyn — Sylva, or a Discourse of Forest TreesSuzanne Simard — Finding the Mother TreePeter Wohlleben — The Hidden Life of TreesRichard Powers — The OverstoryOliver Rackham — Trees and Woodland in the British Landscape; The History of the CountrysideTristan Gooley — How to Read a TreeRay Mears — British Woodland: How to Explore the Secret World of Our ForestsRoger Phillips — UK wild plants and fungi photographic guidesChris Holland — Plant of the Week collectionHandy tools referencedOld OS map viewer for historical layersLiDAR overlays for spotting ridge and furrow, pits and platformsListen now

Landing on the Moon is no small feat, but with advanced landing technologies, NASA is SPLICE-ing the way for safe, autonomous landings.

Amber is a Senior Geologist for the Maine Geological Survey, a state agency in the Department of Agriculture, Conservation, and Forestry that "provides the people and businesses of Maine with essential geologic information about the land where we live and work." Maine has a complex geologic history, and it's made all the harder to study due to the large amount of forest and cover (as opposed to places like New Mexico where the geologic layers are more easily observed).This conversation was recorded in November 2025.  ~~~~~The Maine Science Podcast is a production of the Maine Discovery Museum. It is recorded at Discovery Studios, at the Maine Discovery Museum, in Bangor, ME. The Maine Science Podcast is hosted and executive produced by Kate Dickerson; edited and produced by Scott Loiselle. The Discover Maine theme was composed and performed by Nick Parker. To support our work: https://www.mainediscoverymuseum.org/donate. Find us online:Maine Discovery MuseumMaine Discovery Museum on social media: Facebook Instagram LinkedIn Bluesky YouTubeMaine Science Podcast on social media: Facebook Instagram YouTubeMaine Science Festival on social media: Facebook Instagram LinkedIn YouTube© 2026 Maine Discovery Museum

In this episode, Dr. Jared Henson and Jimbo Robinson welcome OnX Hunt marketing manager and Backwoods University host Lake Pickle. The crew dives into everything from habitat changes in the Mississippi Delta and evolving agricultural pressures to the latest OnX features like LiDAR and collaborative folders. Lake shares his journey from Mississippi kid to Primos videographer to OnX manager, and even unpacks the surprising roots of Santa's flying reindeer. This one blends conservation insight, hunt strategy, mapping tech, family traditions, and plenty of laughs.Listen now: www.ducks.org/DUPodcastSend feedback: DUPodcast@ducks.orgSPONSORS:Purina Pro Plan: The official performance dog food of Ducks UnlimitedWhether you're a seasoned hunter or just getting started, this episode is packed with valuable insights into the world of waterfowl hunting and conservation.Bird Dog Whiskey and Cocktails:Whether you're winding down with your best friend, or celebrating with your favorite crew, Bird Dog brings award-winning flavor to every moment. Enjoy responsibly.

Michael speaks with Craig Singleton, China Program Senior Director and Senior Fellow at the Foundation for Defense of Democracies, about the new frontiers of the US-China tech competition. Craig explains China's willingness to weaponize its dominance in rare earth magnets and how that leverage has left US assembly lines vulnerable. He also explores the high-stakes debate over semiconductor export controls, including a controversial profit-sharing deal for NVIDIA's H20 chips with the US government. Finally, Craig discusses the Chinese "five lever playbook" used to dominate critical sectors like polysilicon, LIDAR, and display technologies, warning of "strategic kill switches" in US infrastructure and the emerging national security threat of biotech.

RISKING IT ALL TO DOCK DRAGON WITH THE ISS Colleague Eric Berger. To fund its Mars ambitions, SpaceX needed NASA contracts to deliver cargo to the International Space Station (ISS) using the Dragon spacecraft. Unlike traditional capsules, Dragon integrated propulsion directly into the vehicle to support future reusability. Behind schedule, SpaceX combined two test missions (C2 and C3) into one high-stakes attempt. During the approach, the spacecraft's LIDAR navigation system faltered, forcing NASA flight director Holly Ridings to make a "brave call": she allowed SpaceX to rewrite software on the fly, defying standard mission rules to achieve a successful docking. NUMBER 3 MAY 1953

In this episode, we review papers that stood out to each of us from the 2025 Wild Turkey Symposium.   Resources:   Collier, B. A., & Chamberlain, M. J. (2025). The Notorious PIG: wild pigs are not primary predators of wild turkey nests. Wildlife Society Bulletin, e1618.   Danks, Z. D., et al. (2025). A national standardized wild turkey brood survey: The first 6 years. Wildlife Society Bulletin, e164   Moscicki, D. J., et al. (2025). Multi‐scale evaluation of eastern wild turkey nest‐site selection and nest survival. Wildlife Society Bulletin, e1635.   Ogawa, R., et al. (2025). Is wild turkey habitat selection spatially consistent? A three‐decade meta‐analysis in Mississippi. Wildlife Society Bulletin, e70000.   The Wildlife Society Bulletin - Wild Turkey Symposium   Thogmartin, W. E. (2001). Home-range size and habitat selection of female wild turkeys (Meleagris gallopavo) in Arkansas. The American Midland Naturalist, 145(2), 247-260.   Ulrey, E. E., et al. (2025). Use of LiDAR to examine habitat selection by incubating female wild turkeys in South Carolina. Wildlife Society Bulletin, e1628.   What does wild turkey nesting cover look like? (Video)   Our lab is primarily funded by donations. If you would like to help support our work, please donate here: http://UFgive.to/UFGameLab   Coming Soon: Wild Turkey Manager: Biology, History, & Heritage! Our newest online wild turkey training is launching soon! Be the first to know when our new course launches by signing up here!   Be sure to check out our comprehensive online wild turkey course featuring experts across multiple institutions that specialize in habitat management and population management for wild turkeys. Earn up to 20.5 CFE hours! Enroll Now!    Dr. Marcus Lashley @DrDisturbance, Publications Dr. Will Gulsby @dr_will_gulsby, Publications Turkeys for Tomorrow @turkeysfortomorrow  UF Game Lab @ufgamelab, YouTube   Want to help wild turkey conservation? Please take our quick survey to take part in our research!   Do you have a topic you'd like us to cover? Leave us a review or send us an email at wildturkeyscience@gmail.com!   Watch these podcasts on YouTube   Please help us by taking our (quick) listener survey - Thank you!    Check out the DrDisturbance YouTube channel! DrDisturbance YouTube   Want to help support the podcast? Our friends at Grounded Brand have an option to donate directly to Wild Turkey Science at checkout. Thank you in advance for your support!   Leave a podcast rating for a chance to win free gear!   This podcast is made possible by Turkeys for Tomorrow, a grassroots organization dedicated to the wild turkey. To learn more about TFT, go to turkeysfortomorrow.org.    Music by Artlist.io Produced & edited by Charlotte Nowak  

In this episode, we review papers that stood out to each of us from the 2025 Wild Turkey Symposium.   Resources:   Collier, B. A., & Chamberlain, M. J. (2025). The Notorious PIG: wild pigs are not primary predators of wild turkey nests. Wildlife Society Bulletin, e1618.   Danks, Z. D., et al. (2025). A national standardized wild turkey brood survey: The first 6 years. Wildlife Society Bulletin, e164   Moscicki, D. J., et al. (2025). Multi‐scale evaluation of eastern wild turkey nest‐site selection and nest survival. Wildlife Society Bulletin, e1635.   Ogawa, R., et al. (2025). Is wild turkey habitat selection spatially consistent? A three‐decade meta‐analysis in Mississippi. Wildlife Society Bulletin, e70000.   The Wildlife Society Bulletin - Wild Turkey Symposium   Thogmartin, W. E. (2001). Home-range size and habitat selection of female wild turkeys (Meleagris gallopavo) in Arkansas. The American Midland Naturalist, 145(2), 247-260.   Ulrey, E. E., et al. (2025). Use of LiDAR to examine habitat selection by incubating female wild turkeys in South Carolina. Wildlife Society Bulletin, e1628.   What does wild turkey nesting cover look like? (Video)   Our lab is primarily funded by donations. If you would like to help support our work, please donate here: http://UFgive.to/UFGameLab   Coming Soon: Wild Turkey Manager: Biology, History, & Heritage! Our newest online wild turkey training is launching soon! Be the first to know when our new course launches by signing up here!   Be sure to check out our comprehensive online wild turkey course featuring experts across multiple institutions that specialize in habitat management and population management for wild turkeys. Earn up to 20.5 CFE hours! Enroll Now!    Dr. Marcus Lashley @DrDisturbance, Publications Dr. Will Gulsby @dr_will_gulsby, Publications Turkeys for Tomorrow @turkeysfortomorrow  UF Game Lab @ufgamelab, YouTube   Want to help wild turkey conservation? Please take our quick survey to take part in our research!   Do you have a topic you'd like us to cover? Leave us a review or send us an email at wildturkeyscience@gmail.com!   Watch these podcasts on YouTube   Please help us by taking our (quick) listener survey - Thank you!    Check out the DrDisturbance YouTube channel! DrDisturbance YouTube   Want to help support the podcast? Our friends at Grounded Brand have an option to donate directly to Wild Turkey Science at checkout. Thank you in advance for your support!   Leave a podcast rating for a chance to win free gear!   This podcast is made possible by Turkeys for Tomorrow, a grassroots organization dedicated to the wild turkey. To learn more about TFT, go to turkeysfortomorrow.org.    Music by Artlist.io Produced & edited by Charlotte Nowak  

In this episode, Seb and Preston explore Tesla's FSD 14.2 advancements and their implications for AI-driven autonomy. They also tackle the ethical, societal, and infrastructural challenges of rapid AI development—from brain-inspired computing to nuclear energy's role in supporting AGI. IN THIS EPISODE YOU'LL LEARN: 00:00:00 - Intro 00:01:44 - How Tesla's FSD 14.2 dramatically improved its autonomous driving performance 00:13:42 - The ethical dilemmas and liability concerns around AI decision-making 00:20:27 - Tesla's sensor-only approach versus LiDAR-heavy systems like Waymo 00:27:31- The potential of biologically-inspired artificial neurons 00:30:32 - How brain-computer interfaces could revolutionize AI and prosthetics 00:32:28 - The societal risks of tech-enhanced human capabilities 00:36:26 - How AI image generation tools like Google's Nano Banana Pro are evolving 00:49:37 - Why AI's energy demands are influencing nuclear power policy 01:00:06 - The risks of AI-induced content homogenization and “AI slop” 01:07:22 - Why some are turning to manual trades to escape AI disruption Disclaimer: Slight discrepancies in the timestamps may occur due to podcast platform differences. BOOKS AND RESOURCES Related book: ⁠Lifespan: Why We Age―and Why We Don't Have To⁠. Seb's website: ⁠Seb Bunney - The Qi of Self Sovereignty. ⁠ Seb's book: ⁠⁠The Hidden Cost of Money⁠⁠. X Account: ⁠Seb Bunney⁠. Related⁠⁠⁠⁠⁠⁠⁠⁠ books⁠⁠⁠⁠⁠⁠⁠⁠ mentioned in the podcast. Ad-free episodes on our⁠⁠⁠⁠⁠⁠⁠⁠ Premium Feed⁠⁠⁠⁠⁠⁠⁠⁠. NEW TO THE SHOW? Join the exclusive ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠TIP Mastermind Community⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ to engage in meaningful stock investing discussions with Stig, Clay, Kyle, and the other community members. Follow our official social media accounts: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠X (Twitter)⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ | ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠LinkedIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ | | ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Instagram⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ | ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Facebook⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ | ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠TikTok⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Check out our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Bitcoin Fundamentals Starter Packs⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Browse through all our episodes (complete with transcripts) ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠here⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Try our tool for picking stock winners and managing our portfolios: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠TIP Finance Tool⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Enjoy exclusive perks from our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠favorite Apps and Services⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Get smarter about valuing businesses in just a few minutes each week through our newsletter, ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠The Intrinsic Value Newsletter⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Learn how to better start, manage, and grow your business with the ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠best business podcasts⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. SPONSORS Support our free podcast by supporting our ⁠⁠⁠⁠⁠⁠⁠⁠⁠sponsors⁠⁠⁠⁠⁠⁠⁠⁠⁠: ⁠Simple Mining⁠ ⁠Human Rights Foundation⁠ ⁠Unchained⁠ ⁠HardBlock⁠ ⁠Linkedin Talent Solutions⁠ ⁠Kubera⁠ ⁠Vanta⁠ ⁠reMarkable⁠ ⁠Onramp⁠ ⁠Public.com⁠ ⁠Netsuite⁠ ⁠Shopify⁠ ⁠Abundant Mines⁠ ⁠Horizon⁠ Learn more about your ad choices. Visit megaphone.fm/adchoices Support our show by becoming a premium member! https://theinvestorspodcastnetwork.supportingcast.fm