Podcasts about llm

In this episode, Varun, co-founder of Yoodli, shares insights into how his startup leverages AI to enhance communication skills, from public speaking to enterprise sales training. Tune in to understand how AI can empower humans rather than replace them, and the strategic evolution from consumer to enterprise products.Key Topics:The origin story of Yoodli and its focus on helping people find their voiceTransition from B2C to B2B: What was learned along the wayThe role of storytelling as a meta-skill in a world dominated by AIUsing AI to make communication more authentic and humanHow large organizations like Google and Snowflake are integrating YoodliThe evolution of AI capabilities, from role plays to experiential learningBuilding modular, customizable AI products that adapt to customer needsThe importance of deep integrations and the challenge of SaaS vendor proliferationReal-world growth stats: 900% revenue increase and millions of usersInsights into leadership, authenticity on social media, and the value of vulnerabilityPersonal stories from Sergey Brin's projects and leadership lessons learnedTimestamps: 00:00 – Introduction to Varun and Yoodli's journey 02:01 – Early days of Yoodli: Founding thesis and initial challenges 04:19 – Key lessons about public speaking skills 05:45 – The importance of recording and reviewing oneself 06:25 – Describing Yoodli as “Duolingo for public speaking” 07:25 – The role of storytelling in high-performance communication 08:21 – Building AI to enhance, not replace, human authenticity 09:07 – Judgment as a differentiator in AI-enabled work 10:01 – How Yoodli expanded into enterprise with Google & others 11:24 – Social media as a branding tool for founders 12:38 – The impact of authenticity on LinkedIn and lead generation 14:09 – The Google GTM training case study: How it started 15:07 – Product features for enterprise sales training 16:05 – Impact on sales onboarding and role play automation 17:32 – The future of experiential learning and AI role plays 20:17 – The broader vision for AI in education and training 21:26 – Impressive growth stats and customer insights 22:01 – The technological foundation: Modular AI architectures 23:52 – The influence of LLM improvements on product features 24:46 – The commoditization of AI role plays and experiential learning 25:12 – Building deep, customizable, scalable AI solutions 26:36 – The importance of scale and deep integrations 30:03 – Product differentiation through vertical focus and deep specialization 33:07 – Market challenges: Demand, consolidation, and customer expectations 34:42 – How to find and connect with Varun 35:30 – Sergey Brin's projects, leadership lessons, and human insights 37:36 – Overcoming imposter syndrome: Everyone's learning curve39:01 – Final reflections and looking aheadResources & Links:Varun on LinkedinNataraj on LinkedinTry Yoodli

We begin the episode with the absolutely ingenious and surprising way in which Kepler discovered the laws of planetary motion.People sometimes say that AI will make especially fast progress at scientific discovery because of tight verification loops.But the story of how we discovered the shape of our solar system shows how the verification loop for correct ideas can be decades (or even millennia) long.During this time, what we know today as the better theory can actually make worse predictions.And the reasons it survives this epistemic hell is some mixture of judgment and heuristics that we don't even understand well enough to actually articulate, much less codify into an RL loop. Hope you enjoy!Watch on YouTube; read the transcript.Sponsors- Jane Street loves challenging my audience with different creative puzzles. One of my listeners, Shawn, solved Jane Street's ResNet challenge and posted a great walk-through on X. If you want to try one of these puzzles yourself, there's one live now at janestreet.com/dwarkesh.- Labelbox can get you rubric-based evals, no matter your domain. These rubrics allow you to give your model feedback on all the dimensions you care about, so you can train how it thinks, not just what it thinks. Whatever you're focused on—math, physics, finance, psychology or something else—Labelbox can help. Learn more at labelbox.com/dwarkesh.- Mercury just released a new feature called Insights. Insights summarizes your money in and out, showing you your biggest transactions and calling out anything worth paying attention to. It's a super low-friction way to stay on top of your business. Learn more at mercury.com/insights.Timestamps(00:00:00) – Kepler was a high temperature LLM(00:11:44) – How would we know if there's a new unifying concept within heaps of AI slop?(00:26:10) – The deductive overhang(00:30:31) – Selection bias in reported AI discoveries(00:46:43) – AI makes papers richer and broader, but not deeper(00:53:00) – If AI solves a problem, can humans get understanding out of it?(00:59:20) – We need a semi-formal language for the way that scientists actually talk to each other(01:09:48) – How Terry uses his time(01:17:05) – Human-AI hybrids will dominate math for a lot longer Get full access to Dwarkesh Podcast at www.dwarkesh.com/subscribe

Leon Trotsky’s Revolution Against God and Christ with James Tunney James Tunney, LLM, is an Irish barrister and author of The Mystery of the Trapped Light: Mystical Thoughts in the Dark Age of Scientism plus The Mystical Accord: Sutras to Suit Our Times, Lines for Spiritual Evolution; also TechBondAge: Slavery of the Human Spirit, Human Entrance to Transhumanism: Machine Merger and the End of Humanity, and AI-Govnerveance: Care and Possession in Dustopia. His most recent book is Trotsky vs Jesus: Battle of the AI-Millennium. His website is https://www.jamestunney.com/ James examines Leon Trotsky as a militant atheist whose vision of permanent, worldwide revolution ultimately leads toward technocracy, posthumanism, and spiritual erasure. He contrasts Trotsky's materialist worldview with Jesus Christ, arguing that Christ represents not a political revolution, but a profound spiritual counter-revolution grounded in moral restraint and inner transformation. Tunney traces how Trotskyist ideas persist across left and right ideologies today, shaping modern systems of power, AI governance, and global control. 00:00:00 Introduction: Trotsky, revolution, and spirituality 00:05:03 Trotsky's historical significance and revolutionary methods 00:10:09 Militant atheism as trostky's driving force 00:15:33 Materialism and technocracy as inevitable outcomes 00:18:26 Permanent revolution and global strategy 00:23:12 Infiltration and political subterfuge 00:29:56 Trotsky in literature and modern politics 00:34:45 Jesus as spiritual counter-revolutionary 00:41:15 AI, posthumanism, and modern power structures 01:10:51 Conclusion New Thinking Allowed host, Jeffrey Mishlove, PhD, is author of The Roots of Consciousness, Psi Development Systems, and The PK Man. Between 1986 and 2002 he hosted and co-produced the original Thinking Allowed public television series. He is the recipient of the only doctoral diploma in “parapsychology” ever awarded by an accredited university (University of California, Berkeley, 1980). He is also the Grand Prize winner of the 2021 Bigelow Institute essay competition regarding the best evidence for survival of human consciousness after permanent bodily death. He is Co-Director of Parapsychology Education at the California Institute for Human Science. (Recorded on Tuesday, February 24, 2026) For a complete, updated list with links to all of our videos, see https://newthinkingallowed.com/Listings.htm. Check out the New Thinking Allowed Foundation website at http://www.newthinkingallowed.org. There you will find our incredible, searchable database as well as opportunities to shop and to support our video productions – plus, this is where people can subscribe to our FREE, weekly Newsletter and can download a FREE .pdf copy of our quarterly magazine. To order high-quality, printed copies of our quarterly magazine: NTA-Magazine.MagCloud.com Check out New Thinking Allowed’s AI chatbot. You can create a free account at awakin.ai/open/jeffreymishlove. When you enter the space, you will see that our chatbot is one of several you can interact with. While it is still a work in progress, it has been trained on 1,600 NTA transcripts. It can provide intelligent answers about the contents of our interviews. It’s almost like having a conversation with Jeffrey Mishlove. If you would like to join our team of volunteers, helping to promote the New Thinking Allowed YouTube channel on social media, editing and translating videos, creating short video trailers based on our interviews, helping to upgrade our website, or contributing in other ways (we may not even have thought of), please send an email to friends@newthinkingallowed.com. To download and listen to audio versions of the New Thinking Allowed videos, please visit our new podcast at https://itunes.apple.com/us/podcast/new-thinking-allowed-audio-podcast/id1435178031. Download and read Jeffrey Mishlove’s Grand Prize essay in the Bigelow Institute competition, Beyond the Brain: The Survival of Human Consciousness After Permanent Bodily Death, go to https://www.bigelowinstitute.org/docs/1st.pdf. You can help support our video productions while enjoying a good book. To order a copy of New Thinking Allowed Dialogues: Is There Life After Death? click on https://amzn.to/3LzLA7Y (As an Amazon Associate we earn from qualifying purchases.) To order the second book in the New Thinking Allowed Dialogues series, Russell Targ: Ninety Years of ESP, Remote Viewing, and Timeless Awareness, go to https://amzn.to/4aw2iyr To order a copy of New Thinking Allowed Dialogues: UFOs and UAP – Are We Really Alone?, go to https://amzn.to/3Y0VOVh To order a copy of Charles T. Tart: Seventy Years of Exploring Consciousness and Parapsychology, go to https://amzn.to/4oOUJLn To order Trotsky vs Jesus: Battle of the AI-Millennium by James Tunney, go to https://amzn.to/46v9Ylb To order AI Govnerveance: Care and Possession in Dustopia by James Tunney, go to https://amzn.to/3ZUeC8D

Most brands fail when treating LLM optimization like traditional SEO. Jeff Reine, co-founder at Everything Machines with two decades in enterprise marketing and platform strategy, has developed Everything Cache to make websites readable for LLM crawlers without rebuilding human-facing sites. The discussion covers why LLM bots behave fundamentally differently from Googlebot and the strategic framework for optimizing content for AI systems versus traditional search engines.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

What can a 40-year-old innovation organization teach today's startup ecosystems about staying relevant and driving impact? In this episode, host Elaine Hamm, PhD, is joined by special guest host James Zanewicz, JD, LLM, RTTP, the original voice behind this podcast and now CEO of Connect in San Diego. James shares insights from stepping into leadership at one of the nation's most established innovation organizations, reflecting on Connect's evolution from a hands-on startup support model to a globally recognized ecosystem builder. The conversation explores what it takes to sustain and grow innovation communities over decades, and how lessons from mature ecosystems can inform emerging regions like the Gulf South. In this episode, you'll learn: Why even the most established innovation organizations must continuously evolve to stay impactful. How Connect is positioning itself as a “front door” to help entrepreneurs navigate dense ecosystems. What founders actually need from ecosystem support and how to deliver it in practical, actionable ways. Tune in to learn how leadership, adaptability, and a founder-first mindset can shape the future of innovation ecosystems in both mature and emerging regions. Links: Connect with James Zanewicz, JD, LLM, RTTP, and learn about Connect. Connect with Elaine Hamm, PhD, and learn about Tulane Medicine Business Development and the School of Medicine. Learn more about Cleantech, San Diego Sports Innovators, and Athena. Connect with Guy Kawasaki, MBA, Petra Stegmann, PhD, and Daniela Gama. Check out the book The First 90 Days. Check out our previous episode with Australia's Alita Singer and David Brown. Connect with Ian McLachlan, BIO from the BAYOU producer. Learn more about BIO from the BAYOU - the podcast. Bio from the Bayou is a podcast that explores biotech innovation, business development, and healthcare outcomes in New Orleans & The Gulf South, connecting biotech companies, investors, and key opinion leaders to advance medicine, technology, and startup opportunities in the region.

AI in healthcare only works when it is built for real workflows, real systems, and real change management. In this episode, Ganesh Padmanabhan, Founder and CEO of Autonomize AI, joins Saul Marquez to unpack what it really takes to deploy AI in healthcare operations at scale. Ganesh shares why many point solutions create islands of efficiency instead of fixing end-to-end processes, and why prompting an LLM is the easy part compared to building production-ready AI that integrates into enterprise systems, meets compliance requirements, and earns clinical trust. He also offers a practical playbook for implementation: anticipate the trust gap, design for behavior change, and treat deployment as a true team sport across tech, clinical, ops, and compliance. Tune in to learn how to move beyond AI hype and build operational AI that delivers measurable value in healthcare! Resources: Connect with and follow Ganesh Padmanabhan on LinkedIn. Follow  Autonomize AI on LinkedIn and discover their website. Check out Ganesh's podcast,  Stories in AI.

Nearly nine in ten B2B buyers have adopted generative AI across their buying process. Jeff Reine, co-founder at Everything Machines, brings two decades of enterprise marketing experience and has built Everything Cache, a brand-side infrastructure that makes websites readable for LLM crawlers without rebuilding human-facing sites. He breaks down the shift from search-and-discover to ask-and-answer behavior, explains why measurement alone isn't sufficient for AI-first discovery, and details the infrastructure framework needed when your first audience isn't human anymore.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Johann Schleier-Smith is the Technical Lead for AI at Temporal Technologies, working on reliable infrastructure for production AI systems and long-running agent workflows. Durable Execution and Modern Distributed Systems, Johann Schleier-Smith // MLOps Podcast #364Join the Community: https://go.mlops.community/YTJoinInGet the newsletter: https://go.mlops.community/YTNewsletterMLOps Merch: https://shop.mlops.community/Big shoutout to ⁨ @Temporalio  for the support, and to  @trychroma  for hosting us in their recording studio// AbstractA new paradigm is emerging for building applications that process large volumes of data, run for long periods of time, and interact with their environment. It's called Durable Execution and is replacing traditional data pipelines with a more flexible approach. Durable Execution makes regular code reliable and scalable.In the past, reliability and scalability have come from restricted programming models, like SQL or MapReduce, but with Durable Execution, this is no longer the case. We can now see data pipelines that include document processing workflows, deep research with LLMs, and other complex and LLM-driven agentic patterns expressed at scale with regular Python programs.In this session, we describe Durable Execution and explain how it fits in with agents and LLMs to enable a new class of machine learning applications.// Related Linkshttps://t.mp/hello?utm_source=podcast&utm_medium=sponsorship&utm_campaign=podcast-2026-03-13-mlops&utm_content=mlops-johannhttps://t.mp/vibe?utm_source=podcast&utm_medium=sponsorship&utm_campaign=podcast-2026-03-13-mlops&utm_content=mlops-johannhttps://t.mp/career?utm_source=podcast&utm_medium=sponsorship&utm_campaign=podcast-2026-03-13-mlops&utm_content=mlops-johann ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreJoin our Slack community [https://go.mlops.community/slack]Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)] Sign up for the next meetup: [https://go.mlops.community/register]MLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Johann on LinkedIn: /jssmith/

In this episode, originally aired on Big Technology Podcast, Olivia Moore discusses whether AI startups can compete with the big chatbots, why American sentiment toward AI is so negative, and what she learned from giving LLMs personality tests. She also breaks down where ChatGPT, Claude, and Gemini are diverging, why Open Claw signals a new wave of agentic products, and what makes memory the most underrated feature in consumer AI.   Resources: Follow Olivia Moore on X:  https://x.com/omooretweets Follow Alex Kantrowitz on X:  https://x.com/Kantrowitz List to Big Technology Podcast: https://www.youtube.com/playlist?list=PLADd6sStSis77HKfbf4KCY6SvthfxeUgn   Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The internet is losing its mind over a new spider chart from Anthropic's latest report on the labor market impacts of AI. However, if you're looking at this chart and using it to predict an AI job apocalypse, you are missing the many leadership lessons playing out right in front of us.While the headlines flying around about it can be deceiving, the reality is a much more sobering masterclass in understanding that this viral chart measures tasks, not jobs. While the media focuses on mass layoffs, the real crisis is what happens when companies assume an LLM can replace human capability. The actual data shows a silent hiring freeze at the entry-level and a looming "gray tsunami" of retiring seasoned experts.This week, I'm breaking down some key insights from the Anthropic AI Labor Impact Report, bunker-busting the spider chart nonsense, and breaking down exactly what the data actually says. I'll explain why AI exposure does not equal job elimination, why assuming "observable" usage equates to actual "effectiveness" is an incredibly dangerous trap, and why companies are suddenly waking up to the fact that you cannot replace your early-career talent pipeline with an AI tool.My goal is to move you out of "Spectator Mode" to "Strategic Preparation" by highlighting the greatest opportunities to prepare your organization for what's ahead.​ Unfreezing Early Career Talent: We love to assume AI will handle all the administrivia, leading to a massive freeze on entry-level hiring. I break down why pausing this pipeline creates a massive future leadership gap. You cannot wait for a crisis to decide how to build talent; you must go to your hiring managers now and ask what these junior roles would do to grow if AI actually did cover the gaps.​ Re-engineering Exposed Roles: We casually assume AI is just coming for administrative work, but the most exposed jobs actually belong to your highly paid, highly educated veterans. I share why you must pair early-career folks with seasoned experts to redesign these roles now, before those veterans retire. You need to ask your top performers exactly where AI consistently gets things wrong before they leave with that intellectual capital.​ Auditing AI Effectiveness: We are making sweeping organizational decisions based on vanity metrics like adoption or output volume. I explain why measuring "observable" tasks as successfully automated is a disaster waiting to happen. You must interrogate your current reports to ensure they measure actual business effectiveness, not just an increase in activity. By the end, I hope you see this massive data report not just as another news cycle, but as a mandate for clarity. You cannot simply wait for the market to dictate your talent strategy; you have to define and fortify the organizational structures that will sustain your business when the pressure is on.⸻If this conversation helps you think more clearly about the future we're building, make sure to like, share, and subscribe. You can also support the show by buying me a coffee at https://buymeacoffee.com/christopherlind  And if your organization is wrestling with how to lead responsibly in the AI era, balancing performance, technology, and people, that's the work I do every day through my consulting and coaching. Learn more at https://christopherlind.co⸻Chapters00:00 – Introduction03:00 – Tasks vs. Jobs07:00 – Exposure vs Elimination10:00 – The Premium Paradox16:00 – Thawing The Entry-Level Hiring Freeze20:00 – "Now What"21:00 – Action 1: The "Pipeline Panic" (Unfreeze Early Career Roles)25:00 – Action 2: The "Gray Tsunami" (Re-engineer Exposed Roles)28:00 – Action 3: The "Activity Illusion" (Audit AI Effectiveness)33:00 – Conclusion & Building Your Roadmap#ArtificialIntelligence #Anthropic #FutureOfWork #Leadership #BusinessStrategy #ChristopherLind #FutureFocused #TalentPipeline #OrganizationalDesign #AIAtWork

Bob Treacy started his career as a union steward on the factory floor at GE Aircraft Engines. After earning a BS and MS in Computer Science from Boston University while working and raising a family at the same time, he jumped to software, never looked back, and remains at the cutting edge with Java and AI. Now Principal Software Architect and Data Engineer at Harvard University, he has been writing Java since 1995 and has attended more than 20 JavaOne conferences. So, he's lived much of the entire life of Java.  At JavaOne 2026 this week he'll present work from Harvard's Dataverse project, which uses LLM embeddings and a graph database knowledge graph to recommend metadata categories for research datasets.  The conversation also covers Java's long evolution, his pragmatic view of AI, and his advice to students to make sure they understand full systems and not just be exclusively a coder. Bob Tracey: LinkedIn | Jim Grisanzio: LinkedIn, X/Twitter

Marty sits down with Brian Murray and Paul Itoi to discuss the convergence of AI agents, graph databases as a solution to LLM memory limitations, and Bitcoin's Lightning Network as the native payment rail for the emerging agentic economy. Paul on X: https://x.com/paulitoi Brian on X: https://x.com/murr STACK SATS hat: https://tftcmerch.io/ Our newsletter: https://www.tftc.io/bitcoin-brief/ TFTC Elite (Ad-free & Discord): https://www.tftc.io/#/portal/signup/ Discord: https://discord.gg/VJ2dABShBz Opportunity Cost Extension: https://www.opportunitycost.app/ Shoutout to our sponsors: Bitkey https://bitkey.world/ OPNEXT https://tinyurl.com/tftc2026 Unchained https://unchained.com/tftc/ SLNT https://slnt.com/tftc Salt of the Earth: https://drinksote.com/tftc Join the TFTC Movement: Main YT Channel https://www.youtube.com/c/TFTC21/videos Clips YT Channel https://www.youtube.com/channel/UCUQcW3jxfQfEUS8kqR5pJtQ Website https://tftc.io/ Newsletter tftc.io/bitcoin-brief/ Twitter https://twitter.com/tftc21 Instagram https://www.instagram.com/tftc.io/ Nostr https://primal.net/tftc Follow Marty Bent: Twitter https://twitter.com/martybent Nostr https://primal.net/martybent Newsletter https://tftc.io/martys-bent/ Podcast https://www.tftc.io/tag/podcasts/

Enregistré au One to One Retail E-Commerce 2026, cet épisode s'attaque à un sujet qui change tout : le commerce agentique.Laurent reçoit Alexandre Chaumien, Head of Revenue Southern Europe chez Shopify, pour parler du nouveau type de client qui arrive - les agents IA - et de ce que les marchands doivent faire pour y être prêts.Au programme :L'Universal Commerce Protocol : ce que Shopify, Google et les grands acteurs du paiement construisent ensemble51% de la Gen Z qui commence son achat sur un LLM : les chiffres qui font réfléchirL'agentic storefront : comment rendre ses produits visibles et compréhensibles pour les agentsPourquoi il faut commencer maintenant, même en EuropeUn épisode au cœur de ce qui redéfinit le commerce pour les prochaines années.Et quelques dernières infos à vous partager :Suivez Le Panier sur Instagram @lepanier.podcast !Inscrivez- vous à la newsletter sur lepanier.io pour cartonner en e-comm !Écoutez les épisodes sur Apple Podcasts, Spotify ou encore Podcast Addict Hébergé par Audiomeans. Visitez audiomeans.fr/politique-de-confidentialite pour plus d'informations.

Ricky Ho is a young entrepreneur, already on his second start-up company that is set to dominate the sourcing products arena. Yes, he has Alibaba to outmanoeuvre, but the signs are good, and his AI-inspired model has much to recommend it to SMBs/SMEs around the world.Summary of PodcastIntroductions and backgroundGraham, Kevin, and Ricky (from SourceReady) introduce themselves and provide background on Ricky's career and the inspiration behind starting SourceReady, a platform that uses AI to help businesses find and vet suppliers outside of China.SourceReady's unique approachRicky explains how SourceReady differentiates itself from platforms like Alibaba. They do this by using a data-driven, objective approach to ranking suppliers based on factors like quality, compliance, and customer relationships - rather than just supplier advertising. This helps democratize access to supplier information for small and medium businesses.Addressing supply chain risksThe discussion shifts to the importance of managing supply chain risks. They discuss such as compliance issues, supplier financial health, and reputational risks. Ricky highlights how SourceReady's tools can help businesses proactively screen for and mitigate these risks, which is especially critical for smaller companies that may lack dedicated procurement teams.SourceReady's growth and futureRicky shares his vision for SourceReady to become the go-to platform for businesses of all sizes to discover and vet suppliers globally, beyond just China. He discusses the challenges of changing user behavior and the need to provide a significantly better experience than existing options to drive adoption. The group also touches on the favorable market trends, like supply chain diversification, that could benefit SourceReady.Recap and closing thoughtsRicky provides a brief testimonial on his experience being interviewed, and the group wraps up the discussion, with Graham and Kevin expressing optimism about SourceReady's potential for success.The Next 100 Days Podcast Co-HostsGraham ArrowsmithGraham founded Finely Fettled eleven years ago to help businesses market to affluent and high-net-worth customers. He's the founder of MicroYES, a Partner of MeclabsAI, providing AI Agents, workflows, and Phone-to-agent delivery systems. Now, Graham offers Answer Engine Optimisation so you get found by LLM search and Enterprise-level AI Solutions.Kevin ApplebyKevin specialises in finance transformation and implementing business change. He's the COO of GrowCFO, which provides both community and CPD-accredited training designed to grow the next generation of finance leaders. You can find Kevin on LinkedIn and at kevinappleby.com

You keep hearing about AI, but nobody is telling you how it actually fits into a biotech career or a job search. That changes today.In this episode, Carina sits down with Heather Karner, a bench scientist with a background in RNA biology who works alongside machine learning researchers in the Bay Area. Heather is actively job searching and has quietly become the go-to AI resource for her lab and her network, not because she is a tech expert, but because she started experimenting and never stopped.Together they share the exact AI use cases they are running right now: a personalized daily brief that flagged Gilead and Eli Lilly RNA acquisitions before they hit LinkedIn, a literature review workflow built for scientists, how to use AI as a tireless teacher for coding and lab protocols, AI note taking that surfaced 10 action items from a 10-minute meeting, and how to turn a rambling brain dump into a clear, professional message.

La tertulia semanal en la que repasamos las últimas noticias de la actualidad científica. En el episodio de hoy: Cara A: -Evento cientófilo para ver el eclipse del 12 de Agosto (6:00) -El libro de Gastón: Agujeros negros. De la relatividad general a la información cuántica (14:30) -¿Por qué alucinan los LLM? (28:00) Este episodio continúa en la Cara B. Contertulios: Juan Carlos Gil, Silvana Tapia, Luisa Achaerandio, Gastón Giribet, Francis Villatoro, Héctor Socas Imagen de portada: NASA. Todos los comentarios vertidos durante la tertulia representan únicamente la opinión de quien los hace... y a veces ni eso

La tertulia semanal en la que repasamos las últimas noticias de la actualidad científica. En el episodio de hoy: -¿Por qué alucinan los LLM? (Continuación) (00:00) -Engranajes que no se tocan (25:40) -El mejor mapa de materia oscura hasta la fecha (49:10) -Cortical Labs pone su sistema con neuronas biológicas DishBrain a jugar al Doom (1:11:40) -Señales de los oyentes (1:48:40) Este episodio es continuación de la Cara A. Contertulios: Juan Carlos Gil, Silvana Tapia, Luisa Achaerandio, Gastón Giribet, Francis Villatoro, Héctor Socas Imagen de portada: NASA. Todos los comentarios vertidos durante la tertulia representan únicamente la opinión de quien los hace... y a veces ni eso

Oura Health, the Finnish wearables company that has sold more than 5 million health tracker rings, is betting on women's health with the launch of its first-ever proprietary large language model designed specifically for women.  “We know historically that women have been underrepresented when it comes to a lot of [medical and pharmaceutical] research,” Tanvi Jayaraman, MD, clinical lead of health AI at Oura, told Glossy. “We want to change that narrative when it comes to women's health.” LLMs are the brains behind AI chatbots, including Oura's in-app Advisor chat where users can ask general wellness questions, specifics about their personal health data or in-depth medical questions.  “Women have been searching for answers [about our health and bodies on the internet] for just as long as the research has been done,” she said. “The answers that [women are] looking for are really disparate and scattered. They're on a niche Reddit forum, or they're kind of word-of-mouth, so a lot of [what we learn online is] hypothesis-driven, data-gathering one-offs.” Starting last year, Dr. Jayaraman's team of board-certified clinicians began “training” Oura's new LLM with only the best data and studies available. This is juxtaposed against many other LLMs, which are trained on the internet at large, which can result in hearsay and causality connections being learned as fact, Dr. Jayaraman said.   “[When we're able to] pick and choose the right training data, the right sources, the right guidelines for women's health, then you can start to push away some of that noise [from the internet],” she said. “Of course, we have a long way to go when it comes to the actual research, but you have to start somewhere.”  Dr. Jayaraman represents a new type of physician who bridges medicine, artificial intelligence and product strategy.  After medical school at Stanford, she worked on AI strategy projects at Bain & Company, working for global diagnostics and pharmaceutical companies, then on Apple's clinical team, where she worked on next-gen digital health tools. She joined Oura last year.  Dr. Jayaraman joined the Glossy Beauty Podcast to discuss Oura's new women-focused LLM, the future of AI-powered wellness chatbots and more.

Turbopuffer came out of a reading app.In 2022, Simon was helping his friends at Readwise scale their infra for a highly requested feature: article recommendations and semantic search. Readwise was paying ~$5k/month for their relational database and vector search would cost ~$20k/month making the feature too expensive to ship. In 2023 after mulling over the problem from Readwise, Simon decided he wanted to “build a search engine” which became Turbopuffer.We discuss:• Simon's path: Denmark → Shopify infra for nearly a decade → “angel engineering” across startups like Readwise, Replicate, and Causal → turbopuffer almost accidentally becoming a company • The Readwise origin story: building an early recommendation engine right after the ChatGPT moment, seeing it work, then realizing it would cost ~$30k/month for a company spending ~$5k/month total on infra and getting obsessed with fixing that cost structure • Why turbopuffer is “a search engine for unstructured data”: Simon's belief that models can learn to reason, but can't compress the world's knowledge into a few terabytes of weights, so they need to connect to systems that hold truth in full fidelity • The three ingredients for building a great database company: a new workload, a new storage architecture, and the ability to eventually support every query plan customers will want on their data • The architecture bet behind turbopuffer: going all in on object storage and NVMe, avoiding a traditional consensus layer, and building around the cloud primitives that only became possible in the last few years • Why Simon hated operating Elasticsearch at Shopify: years of painful on-call experience shaped his obsession with simplicity, performance, and eliminating state spread across multiple systems • The Cursor story: launching turbopuffer as a scrappy side project, getting an email from Cursor the next day, flying out after a 4am call, and helping cut Cursor's costs by 95% while fixing their per-user economics • The Notion story: buying dark fiber, tuning TCP windows, and eating cross-cloud costs because Simon refused to compromise on architecture just to close a deal faster • Why AI changes the build-vs-buy equation: it's less about whether a company can build search infra internally, and more about whether they have time especially if an external team can feel like an extension of their own • Why RAG isn't dead: coding companies still rely heavily on search, and Simon sees hybrid retrieval semantic, text, regex, SQL-style patterns becoming more important, not less • How agentic workloads are changing search: the old pattern was one retrieval call up front; the new pattern is one agent firing many parallel queries at once, turning search into a highly concurrent tool call • Why turbopuffer is reducing query pricing: agentic systems are dramatically increasing query volume, and Simon expects retrieval infra to adapt to huge bursts of concurrent search rather than a small number of carefully chosen calls • The philosophy of “playing with open cards”: Simon's habit of being radically honest with investors, including telling Lachy Groom he'd return the money if turbopuffer didn't hit PMF by year-end • The “P99 engineer”: Simon's framework for building a talent-dense company, rejecting by default unless someone on the team feels strongly enough to fight for the candidate —Simon Hørup Eskildsen• LinkedIn: https://www.linkedin.com/in/sirupsen• X: https://x.com/Sirupsen• https://sirupsen.com/aboutturbopuffer• https://turbopuffer.com/Full Video PodTimestamps00:00:00 The PMF promise to Lachy Groom00:00:25 Intro and Simon's background00:02:19 What turbopuffer actually is00:06:26 Shopify, Elasticsearch, and the pain behind the company00:10:07 The Readwise experiment that sparked turbopuffer00:12:00 The insight Simon couldn't stop thinking about00:17:00 S3 consistency, NVMe, and the architecture bet00:20:12 The Notion story: latency, dark fiber, and conviction00:25:03 Build vs. buy in the age of AI00:26:00 The Cursor story: early launch to breakout customer00:29:00 Why code search still matters00:32:00 Search in the age of agents00:34:22 Pricing turbopuffer in the AI era00:38:17 Why Simon chose Lachy Groom00:41:28 Becoming a founder on purpose00:44:00 The “P99 engineer” philosophy00:49:30 Bending software to your will00:51:13 The future of turbopuffer00:57:05 Simon's tea obsession00:59:03 Tea kits, X Live, and P99 LiveTranscriptSimon Hørup Eskildsen: I don't think I've said this publicly before, but I just called Lockey and was like, local Lockie. Like if this doesn't have PMF by the end of the year, like we'll just like return all the money to you. But it's just like, I don't really, we, Justine and I don't wanna work on this unless it's really working.So we want to give it the best shot this year and like we're really gonna go for it. We're gonna hire a bunch of people. We're just gonna be honest with everyone. Like when I don't know how to play a game, I just play with open cards. Lockey was the only person that didn't, that didn't freak out. He was like, I've never heard anyone say that before.Alessio: Hey everyone, welcome to the Leading Space podcast. This is Celesio Pando, Colonel Laz, and I'm joined by Swix, editor of Leading Space.swyx: Hello. Hello, uh, we're still, uh, recording in the Ker studio for the first time. Very excited. And today we are joined by Simon Eski. Of Turbo Farer welcome.Simon Hørup Eskildsen: Thank you so much for having me.swyx: Turbo Farer has like really gone on a huge tear, and I, I do have to mention that like you're one of, you're not my newest member of the Danish AHU Mafia, where like there's a lot of legendary programmers that have come out of it, like, uh, beyond Trotro, Rasmus, lado Berg and the V eight team and, and Google Maps team.Uh, you're mostly a Canadian now, but isn't that interesting? There's so many, so much like strong Danish presence.Simon Hørup Eskildsen: Yeah, I was writing a post, um, not that long ago about sort of the influences. So I grew up in Denmark, right? I left, I left when, when I was 18 to go to Canada to, to work at Shopify. Um, and so I, like, I've, I would still say that I feel more Danish than, than Canadian.This is also the weird accent. I can't say th because it, this is like, I don't, you know, my wife is also Canadian, um, and I think. I think like one of the things in, in Denmark is just like, there's just such a ruthless pragmatism and there's also a big focus on just aesthetics. Like, they're like very, people really care about like where, what things look like.Um, and like Canada has a lot of attributes, US has, has a lot of attributes, but I think there's been lots of the great things to carry. I don't know what's in the water in Ahu though. Um, and I don't know that I could be considered part of the Mafi mafia quite yet, uh, compared to the phenomenal individuals we just mentioned.Barra OV is also, uh, Danish Canadian. Okay. Yeah. I don't know where he lives now, but, and he's the PHP.swyx: Yeah. And obviously Toby German, but moved to Canada as well. Yes. Like this is like import that, uh, that, that is an interesting, um, talent move.Alessio: I think. I would love to get from you. Definition of Turbo puffer, because I think you could be a Vector db, which is maybe a bad word now in some circles, you could be a search engine.It's like, let, let's just start there and then we'll maybe run through the history of how you got to this point.Simon Hørup Eskildsen: For sure. Yeah. So Turbo Puffer is at this point in time, a search engine, right? We do full text search and we do vector search, and that's really what we're specialized in. If you're trying to do much more than that, like then this might not be the right place yet, but Turbo Buffer is all about search.The other way that I think about it is that we can take all of the world's knowledge, all of the exabytes and exabytes of data that there is, and we can use those tokens to train a model, but we can't compress all of that into a few terabytes of weights, right? Compress into a few terabytes of weights, how to reason with the world, how to make sense of the knowledge.But we have to somehow connect it to something externally that actually holds that like in full fidelity and truth. Um, and that's the thing that we intend to become. Right? That's like a very holier than now kind of phrasing, right? But being the search engine for unstructured, unstructured data is the focus of turbo puffer at this point in time.Alessio: And let's break down. So people might say, well, didn't Elasticsearch already do this? And then some other people might say, is this search on my data, is this like closer to rag than to like a xr, like a public search thing? Like how, how do you segment like the different types of search?Simon Hørup Eskildsen: The way that I generally think about this is like, there's a lot of database companies and I think if you wanna build a really big database company, sort of, you need a couple of ingredients to be in the air.We don't, which only happens roughly every 15 years. You need a new workload. You basically need the ambition that every single company on earth is gonna have data in your database. Multiple times you look at a company like Oracle, right? You will, like, I don't think you can find a company on earth with a digital presence that it not, doesn't somehow have some data in an Oracle database.Right? And I think at this point, that's also true for Snowflake and Databricks, right? 15 years later it's, or even more than that, there's not a company on earth that doesn't, in. Or directly is consuming Snowflake or, or Databricks or any of the big analytics databases. Um, and I think we're in that kind of moment now, right?I don't think you're gonna find a company over the next few years that doesn't directly or indirectly, um, have all their data available for, for search and connect it to ai. So you need that new workload, like you need something to be happening where there's a new workload that causes that to happen, and that new workload is connecting very large amounts of data to ai.The second thing you need. The second condition to build a big database company is that you need some new underlying change in the storage architecture that is not possible from the databases that have come before you. If you look at Snowflake and Databricks, right, commoditized, like massive fleet of HDDs, like that was not possible in it.It just wasn't in the air in the nineties, right? So you just didn't, we just didn't build these systems. S3 and and and so on was not around. And I think the architecture that is now possible that wasn't possible 15 years ago is to go all in on NVME SSDs. It requires a particular type of architecture for the database that.It's difficult to retrofit onto the databases that are already there, including the ones you just mentioned. The second thing is to go all in on OIC storage, more so than we could have done 15 years ago. Like we don't have a consensus layer, we don't really have anything. In fact, you could turn off all the servers that Turbo Buffer has, and we would not lose any data because we have all completely all in on OIC storage.And this means that our architecture is just so simple. So that's the second condition, right? First being a new workload. That means that every company on earth, either indirectly or directly, is using your database. Second being, there's some new storage architecture. That means that the, the companies that have come before you can do what you're doing.I think the third thing you need to do to build a big database company is that over time you have to implement more or less every Cory plan on the data. What that means is that you. You can't just get stuck in, like, this is the one thing that a database does. It has to be ever evolving because when someone has data in the database, they over time expect to be able to ask it more or less every question.So you have to do that to get the storage architecture to the limit of what, what it's capable of. Those are the three conditions.swyx: I just wanted to get a little bit of like the motivation, right? Like, so you left Shopify, you're like principal, engineer, infra guy. Um, you also head of kernel labs, uh, inside of Shopify, right?And then you consulted for read wise and that it kind of gave you that, that idea. I just wanted you to tell that story. Um, maybe I, you've told it before, but, uh, just introduce the, the. People to like the, the new workload, the sort of aha moment for turbo PufferSimon Hørup Eskildsen: For sure. So yeah, I spent almost a decade at Shopify.I was on the infrastructure team, um, from the fairly, fairly early days around 2013. Um, at the time it felt like it was growing so quickly and everything, all the metrics were, you know, doubling year on year compared to the, what companies are contending with today. It's very cute in growth. I feel like lot some companies are seeing that month over month.Um, of course. Shopify compound has been compounding for a very long time now, but I spent a decade doing that and the majority of that was just make sure the site is up today and make sure it's up a year from now. And a lot of that was really just the, um, you know, uh, the Kardashians would drive very, very large amounts of, of data to, to uh, to Shopify as they were rotating through all the merch and building out their businesses.And we just needed to make sure we could handle that. Right. And sometimes these were events, a million requests per second. And so, you know, we, we had our own data centers back in the day and we were moving to the cloud and there was so much sharding work and all of that that we were doing. So I spent a decade just scaling databases ‘cause that's fundamentally what's the most difficult thing to scale about these sites.The database that was the most difficult for me to scale during that time, and that was the most aggravating to be on call for, was elastic search. It was very, very difficult to deal with. And I saw a lot of projects that were just being held back in their ambition by using it.swyx: And I mean, self-hosted.Self-hosted. ‘causeSimon Hørup Eskildsen: it's, yeah, and it commercial, this is like 2015, right? So it's like a very particular vintage. Right. It's probably better at a lot of these things now. Um, it was difficult to contend with and I'm just like, I just think about it. It's an inverted index. It should be good at these kinds of queries and do all of this.And it was, we, we often couldn't get it to do exactly what we needed to do or basically get lucine to do, like expose lucine raw to, to, to what we needed to do. Um, so that was like. Just something that we did on the side and just panic scaled when we needed to, but not a particular focus of mine. So I left, and when I left, I, um, wasn't sure exactly what I wanted to do.I mean, it spent like a decade inside of the same company. I'd like grown up there. I started working there when I was 18.swyx: You only do Rails?Simon Hørup Eskildsen: Yeah. I mean, yeah. Rails. And he's a Rails guy. Uh, love Rails. So good. Um,Alessio: we all wish we could still work in Rails.swyx: I know know. I know, but some, I tried learning Ruby.It's just too much, like too many options to do the same thing. It's, that's my, I I know there's a, there's a way to do it.Simon Hørup Eskildsen: I love it. I don't know that I would use it now, like given cloud code and, and, and cursor and everything, but, um, um, but still it, like if I'm just sitting down and writing a teal code, that's how I think.But anyway, I left and I wasn't, I talked to a couple companies and I was like, I don't. I need to see a little bit more of the world here to know what I'm gonna like focus on next. Um, and so what I decided is like I was gonna, I called it like angel engineering, where I just hopped around in my friend's companies in three months increments and just helped them out with something.Right. And, and just vested a bit of equity and solved some interesting infrastructure problem. So I worked with a bunch of companies at the time, um, read Wise was one of them. Replicate was one of them. Um, causal, I dunno if you've tried this, it's like a, it's a spreadsheet engine Yeah. Where you can do distribution.They sold recently. Yeah. Um, we've been, we used that in fp and a at, um, at Turbo Puffer. Um, so a bunch of companies like this and it was super fun. And so we're the Chachi bt moment happened, I was with. With read Wise for a stint, we were preparing for the reader launch, right? Which is where you, you cue articles and read them later.And I was just getting their Postgres up to snuff, like, which basically boils down to tuning, auto vacuum. So I was doing that and then this happened and we were like, oh, maybe we should build a little recommendation engine and some features to try to hook in the lms. They were not that good yet, but it was clear there was something there.And so I built a small recommendation engine just, okay, let's take the articles that you've recently read, right? Like embed all the articles and then do recommendations. It was good enough that when I ran it on one of the co-founders of Rey's, like I found out that I got articles about, about having a child.I'm like, oh my God, I didn't, I, I didn't know that, that they were having a child. I wasn't sure what to do with that information, but the recommendation engine was good enough that it was suggesting articles, um, about that. And so there was, there was recommendations and uh, it actually worked really well.But this was a company that was spending maybe five grand a month in total on all their infrastructure and. When I did the napkin math on running the embeddings of all the articles, putting them into a vector index, putting it in prod, it's gonna be like 30 grand a month. That just wasn't tenable. Right?Like Read Wise is a proudly bootstrapped company and it's paying 30 grand for infrastructure for one feature versus five. It just wasn't tenable. So sort of in the bucket of this is useful, it's pretty good, but let us, let's return to it when the costs come down.swyx: Did you say it grows by feature? So for five to 30 is by the number of, like, what's the, what's the Scaling factor scale?It scales by the number of articles that you embed.Simon Hørup Eskildsen: It does, but what I meant by that is like five grand for like all of the other, like the Heroku, dinos, Postgres, like all the other, and this then storage is 30. Yeah. And then like 30 grand for one feature. Right. Which is like, what other articles are related to this one.Um, so it was just too much right to, to power everything. Their budget would've been maybe a few thousand dollars, which still would've been a lot. And so we put it in a bucket of, okay, we're gonna do that later. We'll wait, we will wait for the cost to come down. And that haunted me. I couldn't stop thinking about it.I was like, okay, there's clearly some latent demand here. If the cost had been a 10th, we would've shipped it and. This was really the only data point that I had. Right. I didn't, I, I didn't, I didn't go out and talk to anyone else. It was just so I started reading Right. I couldn't, I couldn't help myself.Like I didn't know what like a vector index is. I, I generally barely do about how to generate the vectors. There was a lot of hype about, this is a early 2023. There was a lot of hype about vector databases. There were raising a lot of money and it's like, I really didn't know anything about it. It's like, you know, trying these little models, fine tuning them.Like I was just trying to get sort of a lay of the land. So I just sat down. I have this. A GitHub repository called Napkin Math. And on napkin math, there's just, um, rows of like, oh, this is how much bandwidth. Like this is how many, you know, you can do 25 gigabytes per second on average to dram. You can do, you know, five gigabytes per second of rights to an SSD, blah blah.All of these numbers, right? And S3, how many you could do per, how much bandwidth can you drive per connection? I was just sitting down, I was like, why hasn't anyone build a database where you just put everything on O storage and then you puff it into NVME when you use the data and you puff it into dram if you're, if you're querying it alive, it's just like, this seems fairly obvious and you, the only real downside to that is that if you go all in on o storage, every right will take a couple hundred milliseconds of latency, but from there it's really all upside, right?You do the first go, it takes half a second. And it sort of occurred to me as like, well. The architecture is really good for that. It's really good for AB storage, it's really good for nvm ESSD. It's, well, you just couldn't have done that 10 years ago. Back to what we were talking about before. You really have to build a database where you have as few round trips as possible, right?This is how CPUs work today. It's how NVM E SSDs work. It's how as, um, as three works that you want to have a very large amount of outstanding requests, right? Like basically go to S3, do like that thousand requests to ask for data in one round trip. Wait for that. Get that, like, make a new decision. Do it again, and try to do that maybe a maximum of three times.But no databases were designed that way within NVME as is ds. You can drive like within, you know, within a very low multiple of DRAM bandwidth if you use it that way. And same with S3, right? You can fully max out the network card, which generally is not maxed out. You get very, like, very, very good bandwidth.And, but no one had built a database like that. So I was like, okay, well can't you just, you know, take all the vectors right? And plot them in the proverbial coordinate system. Get the clusters, put a file on S3 called clusters, do json, and then put another file for every cluster, you know, cluster one, do js O cluster two, do js ON you know that like it's two round trips, right?So you get the clusters, you find the closest clusters, and then you download the cluster files like the, the closest end. And you could do this in two round trips.swyx: You were nearest neighbors locally.Simon Hørup Eskildsen: Yes. Yes. And then, and you would build this, this file, right? It's just like ultra simplistic, but it's not a far shot from what the first version of Turbo Buffer was.Why hasn't anyone done thatAlessio: in that moment? From a workload perspective, you're thinking this is gonna be like a read heavy thing because they're doing recommend. Like is the fact that like writes are so expensive now? Oh, with ai you're actually not writing that much.Simon Hørup Eskildsen: At that point I hadn't really thought too much about, well no actually it was always clear to me that there was gonna be a lot of rights because at Shopify, the search clusters were doing, you know, I don't know, tens or hundreds of crew QPS, right?‘cause you just have to have a human sit and type in. But we did, you know, I don't know how many updates there were per second. I'm sure it was in the millions, right into the cluster. So I always knew there was like a 10 to 100 ratio on the read write. In the read wise use case. It's, um, even, even in the read wise use case, there'd probably be a lot fewer reads than writes, right?There's just a lot of churn on the amount of stuff that was going through versus the amount of queries. Um, I wasn't thinking too much about that. I was mostly just thinking about what's the fundamentally cheapest way to build a database in the cloud today using the primitives that you have available.And this is it, right? You just, now you have one machine and you know, let's say you have a terabyte of data in S3, you paid the $200 a month for that, and then maybe five to 10% of that data and needs to be an NV ME SSDs and less than that in dram. Well. You're paying very, very little to inflate the data.swyx: By the way, when you say no one else has done that, uh, would you consider Neon, uh, to be on a similar path in terms of being sort of S3 first and, uh, separating the compute and storage?Simon Hørup Eskildsen: Yeah, I think what I meant with that is, uh, just build a completely new database. I don't know if we were the first, like it was very much, it was, I mean, I, I hadn't, I just looked at the napkin math and was like, this seems really obvious.So I'm sure like a hundred people came up with it at the same time. Like the light bulb and every invention ever. Right. It was just in the air. I think Neon Neon was, was first to it. And they're trying, they're retrofitted onto Postgres, right? And then they built this whole architecture where you have, you have it in memory and then you sort of.You know, m map back to S3. And I think that was very novel at the time to do it for, for all LTP, but I hadn't seen a database that was truly all in, right. Not retrofitting it. The database felt built purely for this no consensus layer. Even using compare and swap on optic storage to do consensus. I hadn't seen anyone go that all in.And I, I mean, there, there, I'm sure there was someone that did that before us. I don't know. I was just looking at the napkin mathswyx: and, and when you say consensus layer, uh, are you strongly relying on S3 Strong consistency? You are. Okay.SoSimon Hørup Eskildsen: that is your consensus layer. It, it is the consistency layer. And I think also, like, this is something that most people don't realize, but S3 only became consistent in December of 2020.swyx: I remember this coming out during COVID and like people were like, oh, like, it was like, uh, it was just like a free upgrade.Simon Hørup Eskildsen: Yeah.swyx: They were just, they just announced it. We saw consistency guys and like, okay, cool.Simon Hørup Eskildsen: And I'm sure that they just, they probably had it in prod for a while and they're just like, it's done right.And people were like, okay, cool. But. That's a big moment, right? Like nv, ME SSDs, were also not in the cloud until around 2017, right? So you just sort of had like 2017 nv, ME SSDs, and people were like, okay, cool. There's like one skew that does this, whatever, right? Takes a few years. And then the second thing is like S3 becomes consistent in 2020.So now it means you don't have to have this like big foundation DB or like zookeeper or whatever sitting there contending with the keys, which is how. You know, that's what Snowflake and others have do so muchswyx: for goneSimon Hørup Eskildsen: Exactly. Just gone. Right? And so just push to the, you know, whatever, how many hundreds of people they have working on S3 solved and then compare and swap was not in S3 at this point in time,swyx: by the way.Uh, I don't know what that is, so maybe you wanna explain. Yes. Yeah.Simon Hørup Eskildsen: Yes. So, um, what Compare and swap is, is basically, you can imagine that if you have a database, it might be really nice to have a file called metadata json. And metadata JSON could say things like, Hey, these keys are here and this file means that, and there's lots of metadata that you have to operate in the database, right?But that's the simplest way to do it. So now you have might, you might have a lot of servers that wanna change the metadata. They might have written a file and want the metadata to contain that file. But you have a hundred nodes that are trying to contend with this metadata that JSON well, what compare and Swap allows you to do is basically just you download the file, you make the modifications, and then you write it only if it hasn't changed.While you did the modification and if not you retry. Right? Should just have this retry loops. Now you can imagine if you have a hundred nodes doing that, it's gonna be really slow, but it will converge over time. That primitive was not available in S3. It wasn't available in S3 until late 2024, but it was available in GCP.The real story of this is certainly not that I sat down and like bake brained it. I was like, okay, we're gonna start on GCS S3 is gonna get it later. Like it was really not that we started, we got really lucky, like we started on GCP and we started on GCP because tur um, Shopify ran on GCP. And so that was the platform I was most available with.Right. Um, and I knew the Canadian team there ‘cause I'd worked with them at Shopify and so it was natural for us to start there. And so when we started building the database, we're like, oh yeah, we have to build a, we really thought we had to build a consensus layer, like have a zookeeper or something to do this.But then we discovered the compare and swap. It's like, oh, we can kick the can. Like we'll just do metadata r json and just, it's fine. It's probably fine. Um, and we just kept kicking the can until we had very, very strong conviction in the idea. Um, and then we kind of just hinged the company on the fact that S3 probably was gonna get this, it started getting really painful in like mid 2024.‘cause we were closing deals with, um, um, notion actually that was running in AWS and we're like, trust us. You, you really want us to run this in GCP? And they're like, no, I don't know about that. Like, we're running everything in AWS and the latency across the cloud were so big and we had so much conviction that we bought like, you know, dark fiber between the AWS regions in, in Oregon, like in the InterExchange and GCP is like, we've never seen a startup like do like, what's going on here?And we're just like, no, we don't wanna do this. We were tuning like TCP windows, like everything to get the latency down ‘cause we had so high conviction in not doing like a, a metadata layer on S3. So those were the three conditions, right? Compare and swap. To do metadata, which wasn't in S3 until late 2024 S3 being consistent, which didn't happen until December, 2020.Uh, 2020. And then NVMe ssd, which didn't end in the cloud until 2017.swyx: I mean, in some ways, like a very big like cloud success story that like you were able to like, uh, put this all together, but also doing things like doing, uh, bind our favor. That that actually is something I've never heard.Simon Hørup Eskildsen: I mean, it's very common when you're a big company, right?You're like connecting your own like data center or whatever. But it's like, it was uniquely just a pain with notion because the, um, the org, like most of the, like if you're buying in Ashburn, Virginia, right? Like US East, the Google, like the GCP and, and AWS data centers are like within a millisecond on, on each other, on the public exchanges.But in Oregon uniquely, the GCP data center sits like a couple hundred kilometers, like east of Portland and the AWS region sits in Portland, but the network exchange they go through is through Seattle. So it's like a full, like 14 milliseconds or something like that. And so anyway, yeah. It's, it's, so we were like, okay, we can't, we have to go through an exchange in Portland.Yeah. Andswyx: you'd rather do this than like run your zookeeper and likeSimon Hørup Eskildsen: Yes. Way rather. It doesn't have state, I don't want state and two systems. Um, and I think all that is just informed by Justine, my co-founder and I had just been on call for so long. And the worst outages are the ones where you have state in multiple places that's not syncing up.So it really came from, from a a, like just a, a very pure source of pain, of just imagining what we would be Okay. Being woken up at 3:00 AM about and having something in zookeeper was not one of them.swyx: You, you're talking to like a notion or something. Do they care or do they just, theySimon Hørup Eskildsen: just, they care about latency.swyx: They latency cost. That's it.Simon Hørup Eskildsen: They just cared about latency. Right. And we just absorbed the cost. We're just like, we have high conviction in this. At some point we can move them to AWS. Right. And so we just, we, we'll buy the fiber, it doesn't matter. Right. Um, and it's like $5,000. Usually when you buy fiber, you buy like multiple lines.And we're like, we can only afford one, but we will just test it that when it goes over the public internet, it's like super smooth. And so we did a lot of, anyway, it's, yeah, it was, that's cool.Alessio: You can imagine talking to the GCP rep and it's like, no, we're gonna buy, because we know we're gonna turn, we're gonna turn from you guys and go to AWS in like six months.But in the meantime we'll do this. It'sSimon Hørup Eskildsen: a, I mean, like they, you know, this workload still runs on GCP for what it's worth. Right? ‘cause it's so, it was just, it was so reliable. So it was never about moving off GCP, it was just about honesty. It was just about giving notion the latency that they deserved.Right. Um, and we didn't want ‘em to have to care about any of this. We also, they were like, oh, egress is gonna be bad. It was like, okay, screw it. Like we're just gonna like vvc, VPC peer with you and AWS we'll eat the cost. Yeah. Whatever needs to be done.Alessio: And what were the actual workloads? Because I think when you think about ai, it's like 14 milliseconds.It's like really doesn't really matter in the scheme of like a model generation.Simon Hørup Eskildsen: Yeah. We were told the latency, right. That we had to beat. Oh, right. So, so we're just looking at the traces. Right. And then sort of like hand draw, like, you know, kind of like looking at the trace and then thinking what are the other extensions of the trace?Right. And there's a lot more to it because it's also when you have, if you have 14 versus seven milliseconds, right. You can fit in another round trip. So we had to tune TCP to try to send as much data in every round trip, prewarm all the connections. And there was, there's a lot of things that compound from having these kinds of round trips, but in the grand scheme it was just like, well, we have to beat the latency of whatever we're up against.swyx: Which is like they, I mean, notion is a database company. They could have done this themselves. They, they do lots of database engineering themselves. How do you even get in the door? Like Yeah, just like talk through that kind of.Simon Hørup Eskildsen: Last time I was in San Francisco, I was talking to one of the engineers actually, who, who was one of our champions, um, at, AT Notion.And they were, they were just trying to make sure that the, you know, per user cost matched the economics that they needed. You know, Uhhuh like, it's like the way I think about, it's like I have to earn a return on whatever the clouds charge me and then my customers have to earn a return on that. And it's like very simple, right?And so there has to be gross margin all the way up and that's how you build the product. And so then our customers have to make the right set of trade off the turbo Puffer makes, and if they're happy with that, that's great.swyx: Do you feel like you're competing with build internally versus buy or buy versus buy?Simon Hørup Eskildsen: Yeah, so, sorry, this was all to build up to your question. So one of the notion engineers told me that they'd sat and probably on a napkin, like drawn out like, why hasn't anyone built this? And then they saw terrible. It was like, well, it literally that. So, and I think AI has also changed the buy versus build equation in terms of, it's not really about can we build it, it's about do we have time to build it?I think they like, I think they felt like, okay, if this is a team that can do that and they, they feel enough like an extension of our team, well then we can go a lot faster, which would be very, very good for them. And I mean, they put us through the, through the test, right? Like we had some very, very long nights to to, to do that POC.And they were really our biggest, our second big customer off the cursor, which also was a lot of late nights. Right.swyx: Yeah. That, I mean, should we go into that story? The, the, the sort of Chris's story, like a lot, um, they credit you a lot for. Working very closely with them. So I just wanna hear, I've heard this, uh, story from Sole's point of view, but like, I'm curious what, what it looks like from your side.Simon Hørup Eskildsen: I actually haven't heard it from Sole's point of view, so maybe you can now cross reference it. The way that I remember it was that, um, the day after we launched, which was just, you know, I'd worked the whole summer on, on the first version. Justine wasn't part of it yet. ‘cause I just, I didn't tell anyone that summer that I was working on this.I was just locked in on building it because it's very easy otherwise to confuse talking about something to actually doing it. And so I was just like, I'm not gonna do that. I'm just gonna do the thing. I launched it and at this point turbo puffer is like a rust binary running on a single eight core machine in a T Marks instance.And me deploying it was like looking at the request log and then like command seeing it or like control seeing it to just like, okay, there's no request. Let's upgrade the binary. Like it was like literally the, the, the, the scrappiest thing. You could imagine it was on purpose because just like at Shopify, we did that all the time.Like, we like move, like we ran things in tux all the time to begin with. Before something had like, at least the inkling of PMF, it was like, okay, is anyone gonna hear about this? Um, and one of the cursor co-founders Arvid reached out and he just, you know, the, the cursor team are like all I-O-I-I-M-O like, um, contenders, right?So they just speak in bullet points and, and facts. It was like this amazing email exchange just of, this is how many QPS we have, this is what we're paying, this is where we're going, blah, blah, blah. And so we're just conversing in bullet points. And I tried to get a call with them a few times, but they were, so, they were like really writing the PMF bowl here, just like late 2023.And one time Swally emails me at like five. What was it like 4:00 AM Pacific time saying like, Hey, are you open for a call now? And I'm on the East coast and I, it was like 7:00 AM I was like, yeah, great, sure, whatever. Um, and we just started talking and something. Then I didn't know anything about sales.It was something that just comp compelled me. I have to go see this team. Like, there's something here. So I, I went to San Francisco and I went to their office and the way that I remember it is that Postgres was down when I showed up at the office. Did SW tell you this? No. Okay. So Postgres was down and so it's like they were distracting with that.And I was trying my best to see if I could, if I could help in any way. Like I knew a little bit about databases back to tuning, auto vacuum. It was like, I think you have to tune out a vacuum. Um, and so we, we talked about that and then, um, that evening just talked about like what would it look like, what would it look like to work with us?And I just said. Look like we're all in, like we will just do what we'll do whatever, whatever you tell us, right? They migrated everything over the next like week or two, and we reduced their cost by 95%, which I think like kind of fixed their per user economics. Um, and it solved a lot of other things. And we were just, Justine, this is also when I asked Justine to come on as my co-founder, she was the best engineer, um, that I ever worked with at Shopify.She lived two blocks away and we were just, okay, we're just gonna get this done. Um, and we did, and so we helped them migrate and we just worked like hell over the next like month or two to make sure that we were never an issue. And that was, that was the cursor story. Yeah.swyx: And, and is code a different workload than normal text?I, I don't know. Is is it just text? Is it the same thing?Simon Hørup Eskildsen: Yeah, so cursor's workload is basically, they, um, they will embed the entire code base, right? So they, they will like chunk it up in whatever they would, they do. They have their own embedding model, um, which they've been public about. Um, and they find that on, on, on their evals.It. There's one of their evals where it's like a 25% improvement on a very particular workload. They have a bunch of blog posts about it. Um, I think it works best on larger code basis, but they've trained their own embedding model to do this. Um, and so you'll see it if you use the cursor agent, it will do searches.And they've also been public around, um, how they've, I think they post trained their model to be very good at semantic search as well. Um, and that's, that's how they use it. And so it's very good at, like, can you find me on the code that's similar to this, or code that does this? And just in, in this queries, they also use GR to supplement it.swyx: Yeah.Simon Hørup Eskildsen: Um, of courseswyx: it's been a big topic of discussion like, is rag dead because gr you know,Simon Hørup Eskildsen: and I mean like, I just, we, we see lots of demand from the coding company to ethicsswyx: search in every part. Yes.Simon Hørup Eskildsen: Uh, we, we, we see demand. And so, I mean, I'm. I like case studies. I don't like, like just doing like thought pieces on this is where it's going.And like trying to be all macroeconomic about ai, that's has turned out to be a giant waste of time because no one can really predict any of this. So I just collect case studies and I mean, cursor has done a great job talking about what they're doing and I hope some of the other coding labs that use Turbo Puffer will do the same.Um, but it does seem to make a difference for particular queries. Um, I mean we can also do text, we can also do RegX, but I should also say that cursors like security posture into Tur Puffer is exceptional, right? They have their own embedding model, which makes it very difficult to reverse engineer. They obfuscate the file paths.They like you. It's very difficult to learn anything about a code base by looking at it. And the other thing they do too is that for their customers, they encrypt it with their encryption keys in turbo puffer's bucket. Um, so it's, it's, it's really, really well designed.swyx: And so this is like extra stuff they did to work with you because you are not part of Cursor.Exactly like, and this is just best practice when working in any database, not just you guys. Okay. Yeah, that makes sense. Yeah. I think for me, like the, the, the learning is kind of like you, like all workloads are hybrid. Like, you know, uh, like you, you want the semantic, you want the text, you want the RegX, you want sql.I dunno. Um, but like, it's silly to like be all in on like one particularly query pattern.Simon Hørup Eskildsen: I think, like I really like the way that, um, um, that swally at cursor talks about it, which is, um, I'm gonna butcher it here. Um, and you know, I'm a, I'm a database scalability person. I'm not a, I, I dunno anything about training models other than, um, what the internet tells me and what.The way he describes is that this is just like cash compute, right? It's like you have a point in time where you're looking at some particular context and focused on some chunk and you say, this is the layer of the neural net at this point in time. That seems fundamentally really useful to do cash compute like that.And, um, how the value of that will change over time. I'm, I'm not sure, but there seems to be a lot of value in that.Alessio: Maybe talk a bit about the evolution of the workload, because even like search, like maybe two years ago it was like one search at the start of like an LLM query to build the context. Now you have a gentech search, however you wanna call it, where like the model is both writing and changing the code and it's searching it again later.Yeah. What are maybe some of the new types of workloads or like changes you've had to make to your architecture for it?Simon Hørup Eskildsen: I think you're right. When I think of rag, I think of, Hey, there's an 8,000 token, uh, context window and you better make it count. Um, and search was a way to do that now. Everything is moving towards the, just let the agent do its thing.Right? And so back to the thing before, right? The LLM is very good at reasoning with the data, and so we're just the tool call, right? And that's increasingly what we see our customers doing. Um, what we're seeing more demand from, from our customers now is to do a lot of concurrency, right? Like Notion does a ridiculous amount of queries in every round trip just because they can't.And I'm also now, when I use the cursor agent, I also see them doing more concurrency than I've ever seen before. So a bit similar to how we designed a database to drive as much concurrency in every round trip as possible. That's also what the agents are doing. So that's new. It means just an enormous amount of queries all at once to the dataset while it's warm in as few turns as possible.swyx: Can I clarify one thing on that?Simon Hørup Eskildsen: Yes.swyx: Is it, are they batching multiple users or one user is driving multiple,Simon Hørup Eskildsen: one user driving multiple, one agent driving.swyx: It's parallel searching a bunch of things.Simon Hørup Eskildsen: Exactly.swyx: Yeah. Yeah, exactly. So yeah, the clinician also did, did this for the fast context thing, like eight parallel at once.Simon Hørup Eskildsen: Yes.swyx: And, and like an interesting problem is, well, how do you make sure you have enough diversity so you're not making the the same request eight times?Simon Hørup Eskildsen: And I think like that's probably also where the hybrid comes in, where. That's another way to diversify. It's a completely different way to, to do the search.That's a big change, right? So before it was really just like one call and then, you know, the LLM took however many seconds to return, but now we just see an enormous amount of queries. So the, um, we just see more queries. So we've like tried to reduce query, we've reduced query pricing. Um, this is probably the first time actually I'm saying that, but the query pricing is being reduced, like five x.Um, and we'll probably try to reduce it even more to accommodate some of these workloads of just doing very large amounts of queries. Um, that's one thing that's changed. I think the right, the right ratio is still very high, right? Like there's still a, an enormous amount of rights per read, but we're starting probably to see that change if people really lean into this pattern.Alessio: Can we talk a little bit about the pricing? I'm curious, uh, because traditionally a database would charge on storage, but now you have the token generation that is so expensive, where like the actual. Value of like a good search query is like much higher because they're like saving inference time down the line.How do you structure that as like, what are people receptive to on the other side too?Simon Hørup Eskildsen: Yeah. I, the, the turbo puffer pricing in the beginning was just very simple. The pricing on these on for search engines before Turbo Puffer was very server full, right? It was like, here's the vm, here's the per hour cost, right?Great. And I just sat down with like a piece of paper and said like, if Turbo Puffer was like really good, this is probably what it would cost with a little bit of margin. And that was the first pricing of Turbo Puffer. And I just like sat down and I was like, okay, like this is like probably the storage amp, but whenever on a piece of paper I, it was vibe pricing.It was very vibe price, and I got it wrong. Oh. Um, well I didn't get it wrong, but like Turbo Puffer wasn't at the first principle pricing, right? So when Cursor came on Turbo Puffer, it was like. Like, I didn't know any VCs. I didn't know, like I was just like, I don't know, I didn't know anything about raising money or anything like that.I just saw that my GCP bill was, was high, was a lot higher than the cursor bill. So Justine and I was just like, well, we have to optimize it. Um, and I mean, to the chagrin now of, of it, of, of the VCs, it now means that we're profitable because we've had so much pricing pressure in the beginning. Because it was running on my credit card and Justine and I had spent like, like tens of thousands of dollars on like compute bills and like spinning off the company and like very like, like bad Canadian lawyers and like things like to like get all of this done because we just like, we didn't know.Right. If you're like steeped in San Francisco, you're just like, you just know. Okay. Like you go out, raise a pre-seed round. I, I never heard a word pre-seed at this point in time.swyx: When you had Cursor, you had Notion you, you had no funding.Simon Hørup Eskildsen: Um, with Cursor we had no funding. Yeah. Um, by the time we had Notion Locke was, Locke was here.Yeah. So it was really just, we vibe priced it 100% from first Principles, but it wasn't, it, it was not performing at first principles, so we just did everything we could to optimize it in the beginning for that, so that at least we could have like a 5% margin or something. So I wasn't freaking out because Cursor's bill was also going like this as they were growing.And so my liability and my credit limit was like actively like calling my bank. It was like, I need a bigger credit. Like it was, yeah. Anyway, that was the beginning. Yeah. But the pricing was, yeah, like storage rights and query. Right. And the, the pricing we have today is basically just that pricing with duct tape and spit to try to approach like, you know, like a, as a margin on the physical underlying hardware.And we're doing this year, you're gonna see more and more pricing changes from us. Yeah.swyx: And like is how much does stuff like VVC peering matter because you're working in AWS land where egress is charged and all that, you know.Simon Hørup Eskildsen: We probably don't like, we have like an enterprise plan that just has like a base fee because we haven't had time to figure out SKU pricing for all of this.Um, but I mean, yeah, you can run turbo puffer either in SaaS, right? That's what Cursor does. You can run it in a single tenant cluster. So it's just you. That's what Notion does. And then you can run it in, in, in BYOC where everything is inside the customer's VPC, that's what an for example, philanthropic does.swyx: What I'm hearing is that this is probably the best CRO job for somebody who can come in and,Simon Hørup Eskildsen: I mean,swyx: help you with this.Simon Hørup Eskildsen: Um, like Turbo Puffer hired, like, I don't know what, what number this was, but we had a full-time CFO as like the 12th hire or something at Turbo Puffer, um, I think I hear are a lot of comp.I don't know how they do it. Like they have a hundred employees and not a CFO. It's like having a CFO is like a runningswyx: business man. Like, you know,Simon Hørup Eskildsen: it's so good. Yeah, like money Mike, like he just, you know, just handles the money and a lot of the business stuff and so he came in and just hopped with a lot of the operational side of the business.So like C-O-O-C-F-O, like somewhere in between.swyx: Just as quick mention of Lucky, just ‘cause I'm curious, I've met Lock and like, he's obviously a very good investor and now on physical intelligence, um, I call it generalist super angel, right? He invests in everything. Um, and I always wonder like, you know, is there something appealing about focusing on developer tooling, focusing on databases, going like, I've invested for 10 years in databases versus being like a lock where he can maybe like connect you to all the customers that you need.Simon Hørup Eskildsen: This is an excellent question. No, no one's asked me this. Um, why lockey? Because. There was a couple of people that we were talking to at the time and when we were raising, we were almost a little, we were like a bit distressed because one of our, one of our peers had just launched something that was very similar to Turbo Puffer.And someone just gave me the advice at the time of just choose the person where you just feel like you can just pick up the phone and not prepare anything. And just be completely honest, and I don't think I've said this publicly before, but I just called Lockey and was like local Lockie. Like if this doesn't have PMF by the end of the year, like we'll just like return all the money to you.But it's just like, I don't really, we, Justine and I don't wanna work on this unless it's really working. So we want to give it the best shot this year and like we're really gonna go for it. We're gonna hire a bunch of people and we're just gonna be honest with everyone. Like when I don't know how to play a game, I just play with open cards and.Lockey was the only person that didn't, that didn't freak out. He was like, I've never heard anyone say that before. As I said, I didn't even know what a seed or pre-seed round was like before, probably even at this time. So I was just like very honest with him. And I asked him like, Lockie, have you ever have, have you ever invested in database company?He was just like, no. And at the time I was like, am I dumb? Like, but I think there was something that just like really drew me to Lockie. He is so authentic, so honest, like, and there was something just like, I just felt like I could just play like, just say everything openly. And that was, that was, I think that that was like a perfect match at the time, and, and, and honestly still is.He was just like, okay, that's great. This is like the most honest, ridiculous thing I've ever heard anyone say to me. But like that, like that, whyswyx: is this ridiculous? Say competitor launch, this may not work out. It wasSimon Hørup Eskildsen: more just like. If this doesn't work out, I'm gonna close up shop by the end of the mo the year, right?Like it was, I don't know, maybe it's common. I, I don't know. He told me it was uncommon. I don't know. Um, that's why we chose him and he'd been phenomenal. The other people were talking at the, at the time were database experts. Like they, you know, knew a lot about databases and Locke didn't, this turned out to be a phenomenal asset.Right. I like Justine and I know a lot about databases. The people that we hire know a lot about databases. What we needed was just someone who didn't know a lot about databases, didn't pretend to know a lot about databases, and just wanted to help us with candidates and customers. And he did. Yeah. And I have a list, right, of the investors that I have a relationship with, and Lockey has just performed excellent in the number of sub bullets of what we can attribute back to him.Just absolutely incredible. And when people talk about like no ego and just the best thing for the founder, I like, I don't think that anyone, like even my lawyer is like, yeah, Lockey is like the most friendly person you will find.swyx: Okay. This is my most glow recommendation I've ever heard.Alessio: He deserves it.He's very special.swyx: Yeah. Yeah. Yeah. Okay. Amazing.Alessio: Since you mentioned candidates, maybe we can talk about team building, you know, like, especially in sf, it feels like it's just easier to start a company than to join a company. Uh, I'm curious your experience, especially not being n SF full-time and doing something that is maybe, you know, a very low level of detail and technical detail.Simon Hørup Eskildsen: Yeah. So joining versus starting, I never thought that I would be a founder. I would start with it, like Turbo Puffer started as a blog post, and then it became a project and then sort of almost accidentally became a company. And now it feels like it's, it's like becoming a bigger company. That was never the intention.The intentions were very pure. It's just like, why hasn't anyone done this? And it's like, I wanna be the, like, I wanna be the first person to do it. I think some founders have this, like, I could never work for anyone else. I, I really don't feel that way. Like, it's just like, I wanna see this happen. And I wanna see it happen with some people that I really enjoy working with and I wanna have fun doing it and this, this, this has all felt very natural on that, on that sense.So it was never a like join versus versus versus found. It was just dis found me at the right moment.Alessio: Well I think there's an argument for, you should have joined Cursor, right? So I'm curious like how you evaluate it. Okay, I should actually go raise money and make this a company versus like, this is like a company that is like growing like crazy.It's like an interesting technical problem. I should just build it within Cursor and then they don't have to encrypt all this stuff. They don't have to obfuscate things. Like was that on your mind at all orSimon Hørup Eskildsen: before taking the, the small check from Lockie, I did have like a hard like look at myself in the mirror of like, okay, do I really want to do this?And because if I take the money, I really have to do it right. And so the way I almost think about it's like you kind of need to ha like you kind of need to be like fucked up enough to want to go all the way. And that was the conversation where I was like, okay, this is gonna be part of my life's journey to build this company and do it in the best way that I possibly can't.Because if I ask people to join me, ask people to get on the cap table, then I have an ultimate responsibility to give it everything. And I don't, I think some people, it doesn't occur to me that everyone takes it that seriously. And maybe I take it too seriously, I don't know. But that was like a very intentional moment.And so then it was very clear like, okay, I'm gonna do this and I'm gonna give it everything.Alessio: A lot of people don't take it this seriously. But,swyx: uh, let's talk about, you have this concept of the P 99 engineer. Uh, people are 10 x saying, everyone's saying, you know, uh, maybe engineers are out of a job. I don't know.But you definitely see a P 99 engineer, and I just want you to talk about it.Simon Hørup Eskildsen: Yeah, so the P 99 engineer was just a term that we started using internally to talk about candidates and talk about how we wanted to build the company. And you know, like everyone else is, like we want a talent dense company.And I think that's almost become trite at this point. What I credit the cursor founders a lot with is that they just arrived there from first principles of like, we just need a talent dense, um, talent dense team. And I think I've seen some teams that weren't talent dense and like seemed a counterfactual run, which if you've run in been in a large company, you will just see that like it's just logically will happen at a large company.Um, and so that was super important to me and Justine and it's very difficult to maintain. And so we just needed, we needed wording for it. And so I have a document called Traits of the P 99 Engineer, and it's a bullet point list. And I look at that list after every single interview that I do, and in every single recap that we do and every recap we end with.End with, um, some version of I'm gonna reject this candidate completely regardless of what the discourse was, because I wanna see people fight for this person because the default should not be, we're gonna hire this person. The default should be, we're definitely not hiring this person. And you know, if everyone was like, ah, maybe throw a punch, then this is not the right.swyx: Do, do you operate, like if there's one cha there must have at least one champion who's like, yes, I will put my career on, on, on the line for this. You know,Simon Hørup Eskildsen: I think career on the line,swyx: maybe a chair, butSimon Hørup Eskildsen: yeah. You know, like, um, I would say so someone needs to like, have both fists up and be like, I'd fight.Right? Yeah. Yeah. And if one person said, then, okay, let's do it. Right?swyx: Yeah.Simon Hørup Eskildsen: Um. It doesn't have to be absolutely everyone. Right? And like the interviews are always the sign that you're checking for different attributes. And if someone is like knocking it outta the park in every single attribute, that's, that's fairly rare.Um, but that's really important. And so the traits of the P 99 engineer, there's lots of them. There's also the traits of the p like triple nine engineer and the quadruple nine engineer. This is like, it's a long list.swyx: Okay.Simon Hørup Eskildsen: Um, I'll give you some samples, right. Of what we, what we look for. I think that the P 99 engineer has some history of having bent, like their trajectory or something to their will.Right? Some moment where it was just, they just, you know, made the computer do what it needed to do. There's something like that, and it will, it will occur to have them at some point in their career. And, uh. Hopefully multiple times. Right.swyx: Gimme an example of one of your engineers that like,Simon Hørup Eskildsen: I'll give an eng.Uh, so we, we, we launched this thing called A and NV three. Um, we could, we're also, we're working on V four and V five right now, but a and NV three can search a hundred billion vectors with a P 50 of around 40 milliseconds and a p 99 of 200 milliseconds. Um, maybe other people have done this, I'm sure Google and others have done this, but, uh, we haven't seen anyone, um, at least not in like a public consumable SaaS that can do this.And that was an engineer, the chief architect of Turbo Puffer, Nathan, um, who more or less just bent this, the software was not capable of this and he just made it capable for a very particular workload in like a, you know, six to eight week period with the help of a lot of the team. Right. It's been, been, there's numerous of examples of that, like at, at turbo puff, but that's like really bending the software and X 86 to your will.It was incredible to watch. Um. You wanna see some moments like that?swyx: Isn't that triple nine?Simon Hørup Eskildsen: Um, I think Nathan, what's calledAlessio: group nine, that was only nine. I feel like this is too high forSimon Hørup Eskildsen: Nathan. Nathan is, uh, Nathan is like, yeah, there's a lot of nines. Okay. After that p So I think that's one trait. I think another trait is that, uh, the P 99 spends a lot of time looking at maps.Generally it's their preferred ux. They just love looking at maps. You ever seen someone who just like, sits on their phone and just like, scrolls around on a map? Or did you not look at maps A lot? You guys don't look atswyx: maps? I guess I'm not feeling there. I don't know, butSimon Hørup Eskildsen: you just dis What about trains?Do you like trains?swyx: Uh, I mean they, not enough. Okay. This is just like weapon nice. Autism is what I call it. Like, like,Simon Hørup Eskildsen: um, I love looking at maps, like, it's like my preferred UX and just like I, you know, I likeswyx: lotsAlessio: of, of like random places, soswyx: like,youswyx: know.Alessio: Yes. Okay. There you go. So instead of like random places, like how do you explore the maps?Simon Hørup Eskildsen: No, it's, it's just a joke.swyx: It's autism laugh. It's like you are just obsessed by something and you like studying a thing.Simon Hørup Eskildsen: The origin of this was that at some point I read an interview with some IOI gold medalistswyx: Uhhuh,Simon Hørup Eskildsen: and it's like, what do you do in your spare time? I was just like, I like looking at maps.I was like, I feel so seen. Like, I just like love, like swirling out. I was like, oh, Canada is so big. Where's Baffin Island? I don't know. I love it. Yeah. Um, anyway, so the traits of P 99, P 99 is obsessive, right? Like, there's just like, you'll, you'll find traits of that we do an interview at, at, at, at turbo puffer or like multiple interviews that just try to screen for some of these things.Um, so. There's lots of others, but these are the kinds of traits that we look for.swyx: I'll tell you, uh, some people listen for like some of my dere stuff. Uh, I do think about derel as maps. Um, you draw a map for people, uh, maps show you the, uh, what is commonly agreed to be the geographical features of what a boundary is.And it shows also shows you what is not doing. And I, I think a lot of like developer tools, companies try to tell you they can do everything, but like, let's, let's be real. Like you, your, your three landmarks are here, everyone comes here, then here, then here, and you draw a map and, and then you draw a journey through the map.And like that. To me, that's what developer relations looks like. So I do think about things that way.Simon Hørup Eskildsen: I think the P 99 thinks in offs, right? The P 99 is very clear about, you know, hey, turbo puffer, you can't run a high transaction workload on turbo puffer, right? It's like the right latency is a hundred milliseconds.That's a clear trade off. I think the P 99 is very good at articulating the trade offs in every decision. Um. Which is exactly what the map is in your case, right?swyx: Uh, yeah, yeah. My, my, my world. My world.Alessio: How, how do you reconcile some of these things when you're saying you bend the will the computer versus like the trade

Manufacturing Hub is back with Episode 252, where co hosts Vlad Romanov and Dave Griffith break down what an AI survival guide should actually look like for manufacturing and industrial automation professionals. This is not a hype conversation about replacing people with magic software. It is a grounded discussion about what AI tools can do today, where they fail, why context and data quality matter so much, and how industrial teams should think about experimentation without losing sight of real operating constraints.In this episode, Vlad and Dave unpack the evolution many engineers and technical leaders have already felt in real time, from early prompt engineering, to agent based workflows, to MCP servers, skills, context management, and the growing cost of tokens and infrastructure. The conversation moves beyond generic AI commentary and into the reality of plant floor environments, where success depends on process knowledge, data architecture, OT constraints, cybersecurity, governance, and clear business value. One of the strongest themes throughout the episode is that manufacturers cannot skip the hard work of structuring data, understanding workflows, and defining use cases simply because AI tools are moving quickly.Vlad brings a very practical industrial lens to the discussion. Drawing on years of hands on experience across controls, manufacturing systems, plant modernization, and digital transformation, he explains why industrial AI has to start with operational context. A maintenance team, an engineering team, and a quality team do not need the same data, do not ask the same questions, and should not be handed the same AI workflows. That distinction matters. This conversation also highlights why the best industrial AI implementations will likely come from teams that combine domain expertise with strong technical execution, rather than generic AI shops trying to force a solution into environments they do not fully understand.Dave adds an important systems and adoption perspective, especially around cost, scaling, management expectations, and the danger of trying to prompt your way past foundational architecture work. Together, Vlad and Dave explore why manufacturers are interested in AI, why many are afraid of being left behind, and why so many projects still stall once they hit the realities of obsolete equipment, weak data models, fragmented systems, and unclear ownership of information. They also discuss deterministic logic versus LLM behavior, reporting workflows, industrial dashboards, PLC code generation concerns, and the practical question every manufacturer should ask before investing: what problem are we solving, for whom, and what is the measurable return?For those new to Vlad, he is an electrical engineer and manufacturing leader with deep experience across industrial automation, controls, data systems, OT architecture, modernization strategy, and plant operations. Through Joltek, Vlad works with manufacturers on digital transformation, IT OT architecture and integration, modernization planning, operational improvement, and technical workforce enablement. Learn more here:Joltek: https://www.joltek.com IT OT Architecture and Integration: https://www.joltek.com/services/service-details-it-ot-architecture-integrationIf you are a plant leader, controls engineer, systems integrator, OT architect, SCADA or MES practitioner, or simply someone trying to separate useful AI workflows from noise, this episode will give you a much more realistic framework for thinking about industrial AI adoption.Timestamps00:00 Welcome back and why this episode matters01:00 Setting up the industrial AI theme for the coming weeks03:10 From prompt engineering to structured AI workflows05:30 AI agents, parallel workflows, tokens, and context windows09:00 MCP tools, Playwright, and what new integrations unlock16:20 How Vlad researches AI and where useful information actually lives22:00 Real manufacturing problems versus AI in search of a problem29:40 Why industrial data architecture is harder than most people think37:00 OT expertise, workforce enablement, and who should build solutions45:40 Practical advice for manufacturers starting the AI journey50:30 Data governance, hallucinations, infrastructure, and cybersecurity57:20 What looks promising today in reporting, dashboards, and industrial applications

Scott Stevenson is the Co-founder and CEO of Spellbook.Spellbook is an AI copilot for contract review and drafting, essentially “Cursor for lawyers.” They have 4,000 customers in 80 countries, and to my knowledge is the fastest growing AI company in Canada, and the largest company in the world built on a Microsoft Word plugin.Scott has been building in legal AI longer than almost anyone. We talk about why legal software was essentially untouched before LLM's, why the market is so hot right now, if it's sustainable, and how Spellbook navigates product differentiation compared to horizontal AI products like ChatGPT.We talk about why fine-tuning your own models was one of the biggest mistakes early AI companies made, how to build a network effect as a vertical AI product, and Spellbook's philosophy of “Don't sharpen your axe when the chainsaw is coming out tomorrow”.Spellbook spent a few years finding PMF before really taking off in 2022, and Scott shares their playbook for launching over 100 product experiments in three years, how to know when to lean in, and what it's been like scaling Spellbook post-PMF.Thank you to Numeral and Flex for supporting this episode.Try Numeral, the end-to-end platform for sales tax and compliance: https://www.numeral.comSign-up for Flex Elite with code TURNER, get $1,000: https://form.typeform.com/to/Rx9rTjFzTimestamps:(0:30) Spellbook: “Cursor for Contracts”(3:08) Building the world's largest Microsoft Word plugin(14:06) Why legal software was untouched before LLMs(18:32) $30 trillion moves through contracts annually(20:51) Why ChatGPT won't replace vertical tools(25:15) Fine-tuning was the biggest mistake in AI(30:00) Differences between pro and amateur gamers(37:38) Top-down vs. bottoms-up in legal AI(42:27) The long-tail of legal AI software(47:24) Building for models that don't exist yet(51:20) Skating where the puck is going(1:01:35) The legal bill that cost 50% of his bank account(1:09:33) Testing 100 landing pages in 3 years(1:14:06) The moment Spellbook hit PMF(1:19:17) Building new brands for each product experiment(1:23:10) Raising a Series B with a tweet(1:27:41) What Scott learned from Keith Rabois(1:31:16) Scott's favorite new AI toolReferencedSpellbook: https://www.spellbook.legal/Careers at Spellbook: https://www.spellbook.legal/careersPlaying to Win by David Sirlin: https://www.amazon.com/Playing-Win-becoming-David-Sirlin/dp/1413498817Find the Fast Moving Water by NFX: https://www.nfx.com/post/find-the-fast-moving-waterSpellbook's case study with Replit: https://replit.com/customers/spellbookTwin: https://twin.so/Follow ScottTwitter: https://x.com/scottastevensonLinkedIn: https://www.linkedin.com/in/scottasBlog: https://blog.scottstevenson.net/Follow TurnerTwitter: https://twitter.com/TurnerNovakLinkedIn: https://www.linkedin.com/in/turnernovakSubscribe to my newsletter to get every episode + the transcript in your inbox every week: https://www.thespl.it/

The E-Commerce Brand's Guide to Unified Performance Marketing in 2026If your affiliate and influencer teams are still working in separate silos with separate budgets and separate tools, this episode will challenge everything about how you've structured your program. Lauryn Day, Director of Agency Partnerships at Levanta, joins Lee-Ann to explain why the lines between affiliate and creator marketing have blurred beyond the point of separate strategies, what a unified approach actually looks like in practice, and why the brands thriving right now are the ones that have stopped treating these two channels as competitors for budget.This is a practical, no-nonsense conversation about how modern e-commerce brands are scaling creator and affiliate programs across Amazon, Shopify, and Walmart, and what the data tells us about where discovery and conversion are actually happening in 2026.About Lauryn DayLauryn Day is Director of Agency Partnerships at Levanta, the leading affiliate marketing software for marketplace sellers. Having started her career on the brand side before moving into tech, she brings a practical, dual-perspective view to creator and affiliate strategy that is rare in this space.Talking Points Include:Why most affiliate infrastructure was built more than 20 years ago and no longer reflects how brands actually operate, and what a modern platform built for today's e-commerce reality looks likeHow Levanta's one-click integration with Amazon, Shopify, and Walmart collapses what used to be a multi-tool, multi-spreadsheet operation into a single five-minute setupWhat AI is actually doing to the shopper journey, and why the data shows less than 10% of consumers who use AI for product research convert through itThe product sampling tool that replaced a spreadsheet-and-email nightmare and why brands are calling it one of the most immediately impactful features they've adoptedHow creator content is now serving two parts of the buying funnel at once: driving direct discovery and feeding the LLM recommendations that consumers use to shortlist productsThe real cost of over-indexing on bottom-funnel partners like coupon and loyalty sites, and how to build a program that serves the full funnelKey Segments of This Podcast and Where You Can Tune In to Go Direct:[01:46] What Levanta is, how it was built, and why most affiliate infrastructure was already out of date before the creator economy arrived[10:24] How consolidating creator and affiliate data into one platform changes the quality of decisions brands can make, with real client context[22:12] Why treating creator and affiliate as separate strategies is the most common and most costly mistake Lauryn sees, and how to fix it[26:30] Lauryn's three actionable tips for brands getting their programs ready for 2026:Consolidate your tech stack. Unify your creator, affiliate, and marketplace channels under one program so your team can focus on strategy rather than administration.Invest in top-of-funnel creator content. UGC builds brand trust, drives direct discovery, and increasingly feeds the AI recommendations consumers use to shortlist products.Stay nimble and test often. Whether it's trialling higher commissions for specific creators, experimenting with a hybrid compensation model, or testing a new content format, run short tests, look at the data, and be willing to pivot quickly.A huge thank you to Lauryn Day for joining us and sharing her insights on where creator and affiliate marketing is heading. If you want to explore what Levanta can do for your program, you can connect with Lauryn directly on LinkeSend me a text with your questions

Roy is a three-time founder who has cracked the code on enterprise AI. After selling his first company and realizing his second idea was too slow, he pivoted to solving a massive problem: customer service automation.In this episode, Roy breaks down how GetVocal went from zero to $1M ARR in just five months. He reveals the "Context Graph" technology that allows them to beat LLM wrappers, why he believes purely generative AI is useless for business, and how he turned a single deployment into an enterprise-wide contagion.Why You Should ListenHow to hit $1M ARR in 5 months with a single salesperson.Why "Context Graphs" are the secret to building AI that doesn't hallucinate.How to expand from a single agent to 80 agents across the enterprise.The critical difference between Deterministic and Probabilistic AI Why starting with a personal passion project failed, but pivoting to enterprise worked.Keywordsstartup podcast, startup podcast for founders, product market fit, enterprise AI, customer service automation, finding pmf, context graphs, AI agents, B2B sales, Roy Moussa00:00:00 Intro00:02:29 From Engineer to 3-Time Founder00:08:11 The Failed Pivot00:12:49 Solving Sales Efficiency First00:16:06 The Pivot to Customer Service00:18:57 Why Chatbots Failed & The Hybrid AI Solution00:25:43 What is a Context Graph?00:34:46 The "Contagion" Effect: 80 Agents in 8 Weeks00:39:34 Competing with Decagon & The Human-Centric Approach00:41:58 Hitting $1M ARR in 5 MonthsSend me a message to let me know what you think!

George Byczynski is a defense and security expert specializing in Central and Eastern Europe. He is an Adviser to the UK's All-Party Parliamentary Group on Poland and a Chief Operating Officer of Anders de Wiart Associates. A former Adviser to the All-Party Parliamentary Groups on Lithuania and Ukraine and a founder of the British Poles Media Group. He holds an LLM in International Law from the University of Westminster and a BSc in International Politics from Brunel University. He is a member of the New Security Leaders of the Warsaw Security Forum and co-author of the reports Three Seas Initiative and the Opportunities for Global Britain and Financing the Future – How to Attract More Foreign Investors to the Three Seas Region. He was awarded the Commission of National Education Honours (KEN) by the Polish Minister of Education and the “Ambassador of Polish History” state award by the Institute of National Remembrance. Byczynski volunteers for the Royal British Legion and the RAF Museum Charity and serves as an Ace Ambassador of the National Spitfire Project. This lecture examines the critical contributions of the British Polish community and the United Kingdom government to Poland's Solidarity movement during the 1980s. It analyzes the significance of the Polish Solidarity Campaign, Solidarity Working Group and the strategic advocacy by Polish émigrés in briefing British parliamentarians, the imposition of economic and diplomatic sanctions on Poland's communist regime following the introduction of martial law in December 1981, the public demonstrations that galvanized support for the Polish cause and the multifaceted approach of British trade unions towards Polish anti-communist movements. The lecture elucidates how these concerted efforts bolstered Poland's struggle for liberty and shaped the broader narrative of international solidarity against the communist oppression. This lecture is part of the 18th Annual Symposium of the Kosciuszko Chair of Polish Studies. The Kościuszko Chair serves as a center for Polish Studies in the broadest sense, including learning, teaching, researching, and writing about Poland's culture, history, heritage, religion, government, economy, and successes in the arts, sciences, and letters, with special emphasis on the achievements of Polish civilization and its relation to other nations, particularly the United States. This year, the 17th annual Kościuszko Chair Conference focuses on the topic of threats and opportunities in the Intermarium. **Learn more about IWP graduate programs: https://www.iwp.edu/academics/graduate-degree-programs/ ***Make a gift to the IWP Kosciuszko Chair of Polish Studies: https://wl.donorperfect.net/weblink/WebLink.aspx?name=E231090&id=4

You've worked hard to build wealth through real estate, but how do you protect your investment when it's time to sell? For most people, the tax law around capital gains is highly complex, and the implications can be downright paralyzing. Beyond significant capital gains tax, you may also owe on depreciable assets, a Medicare surcharge, and state taxes. What are the legal options to minimize the tax burden, and how do they fit into an estate plan?  Jeff sits down with Peter May, JD, LLM, CFP, and Ross Rubin, MBA, of DST Sherpa, LLC, to discuss a powerful—and often overlooked—exit strategy: the Delaware Statutory Trust (DST). As part of a 2004 Internal Revenue Service ruling, a DST lets an investor sell property, defer capital gains taxes through a 1031 Exchange, and transition to passive, hassle-free real estate ownership.  Learn more at https://www.dstsherpa.com/ or contact us to schedule an initial, no-cost consultation. Ross Rubin - ross@transformativenavigation.com  Peter May - peter@transformativenavigation.com.     WHAT YOU NEED TO KNOW (00:00) Episode introduction: Peter May and Ross Rubin of DST Sherpa, LLC (03:16) Creating a real estate exit strategy with a Delaware Statutory Trust (DST) (06:07) Where property investments and estate planning merge (07:49) How DST Sherpa was founded (10:04) The IRS and how a DST works as a 1031 Exchange Replacement Property (14:14) A DST for smaller, accredited investors (16:05) Investment real estate planning for a 1031 Exchange (18:08) Untangling partnership ownership before a sale (19:49) DST Sherpa offers initial consultation at no charge (22:28) The many layers of real estate tax can really add up (23:31) Using a DST when selling a business and its building  (24:56) How to contact Ross and Peter   LINKS AND RESOURCES MENTIONED Bellomo & Associates workshops:https://bellomoassociates.com/workshops/ Life Care Planning The Three Secrets of Estate Planning Nuts & Bolts of Medicaid For more information, call us at (717) 845-5390. Connect with Bellomo & Associates on Social Media Tune in Saturdays at 7:30 a.m. Eastern to WSBA radio: https://www.newstalkwsba.com/ X (formerlyTwitter):https://twitter.com/bellomoassoc YouTube: https://www.youtube.com/user/BellomoAssociates Facebook:https://www.facebook.com/bellomoassociates Instagram:https://www.instagram.com/bellomoassociates/ LinkedIn:https://www.linkedin.com/in/bellomoandassociates WAYS TO WORK WITH JEFFREY BELLOMO Contact Us:https://bellomoassociates.com/contact/ Practice areas:https://bellomoassociates.com/practice-areas/    

This episode is sponsored by tastytrade. Trade stocks, options, futures, and crypto in one platform with low commissions and zero commission on stocks and crypto. Built for traders who think in probabilities, tastytrade offers advanced analytics, risk tools, and an AI-powered Search feature.   Learn more at https://tastytrade.com/ In this episode of the Eye on AI, Craig Smith speaks with Zuzanna Stamirowska about how Pathway is enabling AI systems to work with live, continuously updating data.   Most AI applications rely on static datasets that quickly become outdated. Pathway takes a different approach, allowing developers to build AI systems that process real-time data streams, keeping models, knowledge bases, and AI agents constantly up to date.   Craig and Zuzanna explore why real-time data may be critical for the next generation of LLM applications, RAG systems, and enterprise AI infrastructure, and what it takes to build AI that can operate in a constantly changing world.   Subscribe for more conversations with the researchers and builders shaping the future of AI. Stay Updated: Craig Smith on X: https://x.com/craigss Eye on A.I. on X: https://x.com/EyeOn_AI (00:00) The Core Problem: Why Today's AI Lacks Memory (03:16) Pathway's Mission to Bring Memory Into AI (04:53) Zuzanna's Background in Complexity Science (10:30) Why Transformers Reset Like "Groundhog Day" (14:34) The Brain-Inspired Dragon Hatchling Architecture (23:59) How the Network Learns and Builds Connections (37:38) Performance vs Transformers on Language Tasks (49:37) Productizing the Technology With NVIDIA and AWS (54:23) Can Memory Solve AI Hallucinations?

Send a textReady for a reality check on AI security? We invited Cisco cybersecurity expert Katherine McNamara to dig into where large language models actually break: from prompt injection and over-permissioned plugins to reckless “vibe-coded” apps that leak IDs, photos, and entire backends. The stories are real, the stakes are high, and the fixes are concrete. We trace how AI sprawl mirrors the worst of early IoT—weak defaults, poor isolation, and a stampede to integrate models into billing, HR, and support without guardrails—only this time the blast radius includes your customer data and your legal exposure.We talk through the human factor first. Written policies won't stop someone from pasting a pen test report into a public chatbot. DLP helps, but hybrid work and BYOD stretch defenses thin. Then we move to the core threat model: public and private models are targets; datasets can be poisoned; plugins often ship with admin-level scopes; and a clever prompt can trick an LLM into disclosing chat histories, creating new accounts, or modifying orders. Courts have already treated chatbots as company representatives, binding businesses to their outputs—another reason to treat every integration like an untrusted user with strict least privilege.It's not all doom. Used well, AI gives security operations superpowers: correlating signals across dozens of tools, reducing alert fatigue, and surfacing lateral movement. The path forward is discipline, not denial. Fence models on the network. Prefer read-only to write. Gate plugins behind narrowly scoped APIs. Vet datasets for backdoors. Red-team prompts as seriously as you pen test code. And educate stakeholders with live demos so they see why these controls matter. We also unpack the shaky economics—GPU costs, rising consumer fatigue, hype-fueled projects with little ROI—and why that pressure can erode privacy if teams aren't vigilant.If you're building with LLMs or trying to rein them in, this conversation gives you a practical map: what to allow, what to block, and how to make AI useful without turning your stack into an attack surface. Subscribe, share with a teammate who ships integrations, and drop a review with the one guardrail you'll implement this quarter.Connect with our Guest:https://x.com/kmcnam1https://www.linkedin.com/in/katherinermcnamara/Purchase Chris and Tim's book on AWS Cloud Networking: https://www.amazon.com/Certified-Advanced-Networking-Certification-certification/dp/1835080839/ Check out the Monthly Cloud Networking Newshttps://docs.google.com/document/d/1fkBWCGwXDUX9OfZ9_MvSVup8tJJzJeqrauaE6VPT2b0/Visit our website and subscribe: https://www.cables2clouds.com/Follow us on BlueSky: https://bsky.app/profile/cables2clouds.comFollow us on YouTube: https://www.youtube.com/@cables2clouds/Follow us on TikTok: https://www.tiktok.com/@cables2cloudsMerch Store: https://store.cables2clouds.com/Join the Discord Study group: https://artofneteng.com/iaatj

Baruch Toledano breaks down SimilarWeb's Generative AI Brand Visibility Report. “We measure real user prompts and responses,” and by understanding trending topics and which brands show up in LLM answers, they find that visibility is “highly concentrated.” Less well-known brands that are mentioned structure their content “correctly” for the AI to read, and are highly knowledgeable about specific niches. Baruch explains how metrics are moving from ranking to mentions.======== 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

Robert Cantwell and Luke Yang break down Oracle (ORCL) earnings, with the stock up around 10% after the report. Robert estimates LLM revenue is around $50 billion total and heading quickly to $100 billion. However, Oracle is spending money faster than they're bringing it in, he notes, and he is suspicious around their expense reporting. He argues that CoreWeave (CRWV) is a “pure play” on what Oracle is trying to do. Luke is looking for more clarity on the financing and how it can ramp up its business. He also has concerns around customer concentration.======== 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

Reddit's Chief Operating Officer Jen Wong discusses the impact of gen AI models on its platform and how the company is positioning itself in the era of chatbots and LLMs. Wong sits down with Bloomberg Intelligence's Global Head of Technology Research Mandeep Singh to discuss  the company's ads business and how it plans to leverage LLM search to boost engagement. All the metrics referenced in the episode are as of December 2025.

This is a recap of the top 10 posts on Hacker News on March 10, 2026. This podcast was generated by wondercraft.ai (00:30): Tony Hoare has diedOriginal post: https://news.ycombinator.com/item?id=47324054&utm_source=wondercraft_ai(01:54): Online age-verification tools for child safety are surveilling adultsOriginal post: https://news.ycombinator.com/item?id=47322635&utm_source=wondercraft_ai(03:19): After outages, Amazon to make senior engineers sign off on AI-assisted changesOriginal post: https://news.ycombinator.com/item?id=47323017&utm_source=wondercraft_ai(04:44): Meta acquires MoltbookOriginal post: https://news.ycombinator.com/item?id=47323900&utm_source=wondercraft_ai(06:09): I put my whole life into a single databaseOriginal post: https://news.ycombinator.com/item?id=47321233&utm_source=wondercraft_ai(07:34): Yann LeCun raises $1B to build AI that understands the physical worldOriginal post: https://news.ycombinator.com/item?id=47320600&utm_source=wondercraft_ai(08:59): Redox OS has adopted a Certificate of Origin policy and a strict no-LLM policyOriginal post: https://news.ycombinator.com/item?id=47320661&utm_source=wondercraft_ai(10:24): Show HN: How I topped the HuggingFace open LLM leaderboard on two gaming GPUsOriginal post: https://news.ycombinator.com/item?id=47322887&utm_source=wondercraft_ai(11:49): Two Years of Emacs SoloOriginal post: https://news.ycombinator.com/item?id=47317616&utm_source=wondercraft_ai(13:14): Debian decides not to decide on AI-generated contributionsOriginal post: https://news.ycombinator.com/item?id=47324087&utm_source=wondercraft_aiThis is a third-party project, independent from HN and YC. Text and audio generated using AI, by wondercraft.ai. Create your own studio quality podcast with text as the only input in seconds at app.wondercraft.ai. Issues or feedback? We'd love to hear from you: team@wondercraft.ai

This Week In Startups is made possible by:Northwest Registered Agent - ⁠⁠northwestregisteredagent.com/twist⁠⁠Quo - ⁠⁠quo.com/TWiST⁠⁠Gusto - ⁠⁠Gusto.com/twist⁠⁠Plaud - http://Plaud.ai/twistAthena - https://www.athena.com/jcalToday's show:How long until AI models can improve AI models? Once possible, recursive self-improvement by AI technology could accelerate — forever. Thus far, humans (and their coding agents) are still driving AI progress. But a recent project by AI developer extraordinare Andrej Karpathy, called ‘autoresearcher', is turning heads as it shows that it is possible — in certain contexts — to allow AI agents to run successive coding experiments to improve specific elements of LLM performance. Call it an early demonstration of the future.OpenClaw is exploding in China, while here in the United States, AI is polling somewhere underneath the basement. AI in the United States is about as popular as ICE, which could create a political issue for the technology in the coming elections.Next? Three demos. First, NetXD's Suresh Ramamurthi showed off how he has built OpenClaw functionality to move money, Rohan Arun showed off PhoneClaw automation on Android devices from an AR headset, and Eugene Stuckless gave us a taste of what Eir is building. Our takeaway? OpenClaw is still boring its way into our digital lives, one new skill or tool at a time!GUESTS:Suresh Ramamurthi: ⁠⁠https://x.com/sureshr7⁠⁠Rohan Arun: ⁠⁠https://x.com/Viewforge/⁠⁠Eugene Stuckless: ⁠⁠https://x.com/eugene_eir_inc⁠⁠Timestamps:0:00 — ‘Autoresearcher' and the future of AI improvements6:52 — Why people around the world are flocking to OpenClaw7:57 — Plaud - If your work depends on conversations — interviews, meetings, calls — you need a Plaud NotePin. You can check it out at Plaud.ai/twist and use code TWIST for 10% off!9:46 — Gusto - Check out the online payroll and benefits experts with software built specifically for small business and startups. Try Gusto today and get three months FREE at Gusto.com/twist.12:57 — The changing American social contract20:15 — Quo - Quo (formerly OpenPhone) gives you a clean, modern way to handle every customer call, text, and thread all in one place. Try it free at quo.com/TWIST23:50 — Why China is all-in on AI (and Europe isn't)26:26 — How to keep your job in the AI era28:05 — Northwest Registered Agent - Get more when you start your business with Northwest. In 10 clicks and 10 minutes, you can form your company and walk away with a real business identity — Learn more at www.northwestregisteredagent.com/twist29:38 — Athena - Get $2,000 off your first EA at https://www.athena.com/jcal34:42 — Demo: Suresh Ramamurthi of NetXD42:47 — Demo: Rohan Arun of PhoneClaw47:35 — Why bringing OpenClaw to your smartphone is what's next49:49 — Demo: Eugene Stuckless of Eir56:45 — How can we make smarter, more efficient agents?Subscribe to the TWiST500 newsletter: ⁠⁠https://ticker.thisweekinstartups.com⁠⁠Check out the TWIST500: ⁠⁠https://www.twist500.com⁠⁠Subscribe to This Week in Startups on Apple: ⁠⁠https://rb.gy/v19fcp⁠⁠Follow Lon:X: ⁠⁠https://x.com/lons⁠⁠Follow Alex:X: ⁠⁠https://x.com/alex⁠⁠LinkedIn: ⁠⁠⁠https://www.linkedin.com/in/alexwilhelm⁠⁠Follow Jason:X: ⁠⁠https://twitter.com/Jason⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/in/jasoncalacanis⁠⁠Great TWIST interviews: ⁠⁠Will Guidara,⁠⁠ ⁠⁠Eoghan McCabe⁠⁠, ⁠⁠Steve Huffman⁠⁠, ⁠⁠Brian Chesky⁠⁠, ⁠⁠Bob Moesta,⁠⁠ ⁠⁠Aaron Levie⁠⁠, ⁠⁠Sophia Amoruso⁠⁠, ⁠⁠Reid Hoffman⁠⁠, ⁠⁠Frank Slootman⁠⁠, ⁠⁠Billy McFarland⁠⁠Check out Jason's suite of newsletters: ⁠⁠https://substack.com/@calacanis⁠⁠Follow TWiST:Twitter: ⁠⁠https://twitter.com/TWiStartups⁠⁠YouTube: ⁠⁠https://www.youtube.com/thisweekin⁠⁠Instagram: ⁠⁠https://www.instagram.com/thisweekinstartups⁠⁠TikTok: ⁠⁠https://www.tiktok.com/@thisweekinstartups⁠⁠Substack: ⁠⁠https://twistartups.substack.com

Discord delays their age-gating rollout but legislators are pushing for operating systems including Linux to verify ages, LLM licence laundering might mean the end of copyleft, and how and why you might want to detect Meta’s spy camera glasses. News Getting Global Age Assurance Right: What We Got Wrong and What’s Changing US state laws push age checks into the operating system California introduces age verification law for all operating systems, including Linux and SteamOS — user age verified during OS account setup I have actually read the text of California law CA AB1043 and, honestly, I don’t hate it Do you really think that circumventing these things will always be a simple firmware mod or hardware hack? Relicensing with AI-assisted rewrite – Tuan-Anh Tran Chardet dispute shows how AI will kill software licensing, argues Bruce Perens No right to relicense this project Hide from Meta’s spyglasses with this new Android app Dear Meta Smart Glasses Wearers: You’re Being Watched, Too Automox Turnkey Results Endpoint management tailored to your specific environment. Know the plan. Trust the result. Learn more at www.automox.com Support us on patreon and get an ad-free RSS feed with early episodes sometimes See our contact page for ways to get in touch. RSS: Subscribe to the RSS feeds here

In this episode of Next in Media, I sit down with Sam Garfield, Head of Digital Strategy for CMT Data and AI Platforms at Adobe, to explore how Adobe is quietly becoming the backbone of modern marketing. Sam breaks down how Adobe operates across three core layers: the creative layer (Creative Cloud and Firefly AI), the content supply chain layer (Workfront and asset management), and the data and experience layer (customer data platforms and analytics). Together, these tools form what Sam describes as an operating system for marketers -- a full-stack solution that takes a brand from ideation all the way through activation and measurement. We also dig into the rise of creative intelligence and what it means for brands, agencies, and the future of advertising. Sam unpacks Adobe's Winterberry Group research showing a 23% increase in investment in creative intelligence, and explains why creative can no longer be treated as a fixed cost. We cover how generative AI is accelerating asset production at scale, why agencies are leaning into Adobe's platform rather than building from scratch, and how agentic AI is beginning to appear inside existing workflows. Sam also reveals that traffic to brand sites and publishers is down 40% as LLMs reshape discovery, and shares how Adobe's new LLM Optimizer tool is helping brands regain visibility in a generative search world.   Key Highlights

Join Kyle, Nader, Vibhu, and swyx live at NVIDIA GTC next week!Now that AIE Europe tix are ~sold out, our attention turns to Miami and World's Fair!The definitive AI Accelerator chip company has more than 10xed this AI Summer:And is now a $4.4 trillion megacorp… that is somehow still moving like a startup. We are blessed to have a unique relationship with our first ever NVIDIA guests: Kyle Kranen who gave a great inference keynote at the first World's Fair and is one of the leading architects of NVIDIA Dynamo (a Datacenter scale inference framework supporting SGLang, TRT-LLM, vLLM), and Nader Khalil, a friend of swyx from our days in Celo in The Arena, who has been drawing developers at GTC since before they were even a glimmer in the eye of NVIDIA:Nader discusses how NVIDIA Brev has drastically reduced the barriers to entry for developers to get a top of the line GPU up and running, and Kyle explains NVIDIA Dynamo as a data center scale inference engine that optimizes serving by scaling out, leveraging techniques like prefill/decode disaggregation, scheduling, and Kubernetes-based orchestration, framed around cost, latency, and quality tradeoffs. We also dive into Jensen's “SOL” (Speed of Light) first-principles urgency concept, long-context limits and model/hardware co-design, internal model APIs (https://build.nvidia.com), and upcoming Dynamo and agent sessions at GTC.Full Video pod on YouTubeTimestamps00:00 Agent Security Basics00:39 Podcast Welcome and Guests07:19 Acquisition and DevEx Shift13:48 SOL Culture and Dynamo Setup27:38 Why Scale Out Wins29:02 Scale Up Limits Explained30:24 From Laptop to Multi Node33:07 Cost Quality Latency Tradeoffs38:42 Disaggregation Prefill vs Decode41:05 Kubernetes Scaling with Grove43:20 Context Length and Co Design57:34 Security Meets Agents58:01 Agent Permissions Model59:10 Build Nvidia Inference Gateway01:01:52 Hackathons And Autonomy Dreams01:10:26 Local GPUs And Scaling Inference01:15:31 Long Running Agents And SF ReflectionsTranscriptAgent Security BasicsNader: Agents can do three things. They can access your files, they can access the internet, and then now they can write custom code and execute it. You literally only let an agent do two of those three things. If you can access your files and you can write custom code, you don't want internet access because that's one to see full vulnerability, right?If you have access to internet and your file system, you should know the full scope of what that agent's capable of doing. Otherwise, now we can get injected or something that can happen. And so that's a lot of what we've been thinking about is like, you know, how do we both enable this because it's clearly the future.But then also, you know, what, what are these enforcement points that we can start to like protect?swyx: All right.Podcast Welcome and Guestsswyx: Welcome to the Lean Space podcast in the Chromo studio. Welcome to all the guests here. Uh, we are back with our guest host Viu. Welcome. Good to have you back. And our friends, uh, Netter and Kyle from Nvidia. Welcome.Kyle: Yeah, thanks for having us.swyx: Yeah, thank you. Actually, I don't even know your titles.Uh, I know you're like architect something of Dynamo.Kyle: Yeah. I, I'm one of the engineering leaders [00:01:00] and a architects of Dynamo.swyx: And you're director of something and developers, developer tech.Nader: Yeah.swyx: You're the developers, developers, developers guy at nvidia,Nader: open source agent marketing, brev,swyx: and likeNader: Devrel tools and stuff.swyx: Yeah. BeenNader: the focus.swyx: And we're, we're kind of recording this ahead of Nvidia, GTC, which is coming to town, uh, again, uh, or taking over town, uh, which, uh, which we'll all be at. Um, and we'll talk a little bit about your sessions and stuff. Yeah.Nader: We're super excited for it.GTC Booth Stunt Storiesswyx: One of my favorite memories for Nader, like you always do like marketing stunts and like while you were at Rev, you like had this surfboard that you like, went down to GTC with and like, NA Nvidia apparently, like did so much that they bought you.Like what, what was that like? What was that?Nader: Yeah. Yeah, we, we, um. Our logo was a chaka. We, we, uh, we were always just kind of like trying to keep true to who we were. I think, you know, some stuff, startups, you're like trying to pretend that you're a bigger, more mature company than you are. And it was actually Evan Conrad from SF Compute who was just like, you guys are like previousswyx: guest.Yeah.Nader: Amazing. Oh, really? Amazing. Yeah. He was just like, guys, you're two dudes in the room. Why are you [00:02:00] pretending that you're not? Uh, and so then we were like, okay, let's make the logo a shaka. We brought surfboards to our booth to GTC and the energy was great. Yeah. Some palm trees too. They,Kyle: they actually poked out over like the, the walls so you could, you could see the bread booth.Oh, that's so funny. AndNader: no one else,Kyle: just from very far away.Nader: Oh, so you remember it backKyle: then? Yeah I remember it pre-acquisition. I was like, oh, those guys look cool,Nader: dude. That makes sense. ‘cause uh, we, so we signed up really last minute, and so we had the last booth. It was all the way in the corner. And so I was, I was worried that no one was gonna come.So that's why we had like the palm trees. We really came in with the surfboards. We even had one of our investors bring her dog and then she was just like walking the dog around to try to like, bring energy towards our booth. Yeah.swyx: Steph.Kyle: Yeah. Yeah, she's the best,swyx: you know, as a conference organizer, I love that.Right? Like, it's like everyone who sponsors a conference comes, does their booth. They're like, we are changing the future of ai or something, some generic b******t and like, no, like actually try to stand out, make it fun, right? And people still remember it after three years.Nader: Yeah. Yeah. You know what's so funny?I'll, I'll send, I'll give you this clip if you wanna, if you wanna add it [00:03:00] in, but, uh, my wife was at the time fiance, she was in medical school and she came to help us. ‘cause it was like a big moment for us. And so we, we bought this cricket, it's like a vinyl, like a vinyl, uh, printer. ‘cause like, how else are we gonna label the surfboard?So, we got a surfboard, luckily was able to purchase that on the company card. We got a cricket and it was just like fine tuning for enterprises or something like that, that we put on the. On the surfboard and it's 1:00 AM the day before we go to GTC. She's helping me put these like vinyl stickers on.And she goes, you son of, she's like, if you pull this off, you son of a b***h. And so, uh, right. Pretty much after the acquisition, I stitched that with the mag music acquisition. I sent it to our family group chat. Ohswyx: Yeah. No, well, she, she made a good choice there. Was that like basically the origin story for Launchable is that we, it was, and maybe we should explain what Brev is andNader: Yeah.Yeah. Uh, I mean, brev is just, it's a developer tool that makes it really easy to get a GPU. So we connect a bunch of different GPU sources. So the basics of it is like, how quickly can we SSH you into a G, into a GPU and whenever we would talk to users, they wanted A GPU. They wanted an A 100. And if you go to like any cloud [00:04:00] provisioning page, usually it's like three pages of forms or in the forms somewhere there's a dropdown.And in the dropdown there's some weird code that you know to translate to an A 100. And I remember just thinking like. Every time someone says they want an A 100, like the piece of text that they're telling me that they want is like, stuffed away in the corner. Yeah. And so we were like, what if the biggest piece of text was what the user's asking for?And so when you go to Brev, it's just big GPU chips with the type that you want withswyx: beautiful animations that you worked on pre, like pre you can, like, now you can just prompt it. But back in the day. Yeah. Yeah. Those were handcraft, handcrafted artisanal code.Nader: Yeah. I was actually really proud of that because, uh, it was an, i I made it in Figma.Yeah. And then I found, I was like really struggling to figure out how to turn it from like Figma to react. So what it actually is, is just an SVG and I, I have all the styles and so when you change the chip, whether it's like active or not it changes the SVG code and that somehow like renders like, looks like it's animating, but it, we just had the transition slow, but it's just like the, a JavaScript function to change the like underlying SVG.Yeah. And that was how I ended up like figuring out how to move it from from Figma. But yeah, that's Art Artisan. [00:05:00]Kyle: Speaking of marketing stunts though, he actually used those SVGs. Or kind of use those SVGs to make these cards.Nader: Oh yeah. LikeKyle: a GPU gift card Yes. That he handed out everywhere. That was actually my first impression of thatNader: one.Yeah,swyx: yeah, yeah.Nader: Yeah.swyx: I think I still have one of them.Nader: They look great.Kyle: Yeah.Nader: I have a ton of them still actually in our garage, which just, they don't have labels. We should honestly like bring, bring them back. But, um, I found this old printing press here, actually just around the corner on Ven ness. And it's a third generation San Francisco shop.And so I come in an excited startup founder trying to like, and they just have this crazy old machinery and I'm in awe. ‘cause the the whole building is so physical. Like you're seeing these machines, they have like pedals to like move these saws and whatever. I don't know what this machinery is, but I saw all three generations.Like there's like the grandpa, the father and the son, and the son was like, around my age. Well,swyx: it's like a holy, holy trinity.Nader: It's funny because we, so I just took the same SVG and we just like printed it and it's foil printing, so they make a a, a mold. That's like an inverse of like the A 100 and then they put the foil on it [00:06:00] and then they press it into the paper.And I remember once we got them, he was like, Hey, don't forget about us. You know, I guess like early Apple and Cisco's first business cards were all made there. And so he was like, yeah, we, we get like the startup businesses but then as they mature, they kind of go somewhere else. And so I actually, I think we were talking with marketing about like using them for some, we should go back and make some cards.swyx: Yeah, yeah, yeah. You know, I remember, you know, as a very, very small breadth investor, I was like, why are we spending time like, doing these like stunts for GPUs? Like, you know, I think like as a, you know, typical like cloud hard hardware person, you go into an AWS you pick like T five X xl, whatever, and it's just like from a list and you look at the specs like, why animate this GP?And, and I, I do think like it just shows the level of care that goes throughout birth and Yeah. And now, and also the, and,Nader: and Nvidia. I think that's what the, the thing that struck me most when we first came in was like the amount of passion that everyone has. Like, I think, um, you know, you talk to, you talk to Kyle, you talk to, like, every VP that I've met at Nvidia goes so close to the metal.Like, I remember it was almost a year ago, and like my VP asked me, he's like, Hey, [00:07:00] what's cursor? And like, are you using it? And if so, why? Surprised at this, and he downloaded Cursor and he was asking me to help him like, use it. And I thought that was, uh, or like, just show him what he, you know, why we were using it.And so, the amount of care that I think everyone has and the passion, appreciate, passion and appreciation for the moment. Right. This is a very unique time. So it's really cool to see everyone really like, uh, appreciate that.swyx: Yeah.Acquisition and DevEx Shiftswyx: One thing I wanted to do before we move over to sort of like research topics and, uh, the, the stuff that Kyle's working on is just tell the story of the acquisition, right?Like, not many people have been, been through an acquisition with Nvidia. What's it like? Uh, what, yeah, just anything you'd like to say.Nader: It's a crazy experience. I think, uh, you know, we were the thing that was the most exciting for us was. Our goal was just to make it easier for developers.We wanted to find access to GPUs, make it easier to do that. And then all, oh, actually your question about launchable. So launchable was just make one click exper, like one click deploys for any software on top of the GPU. Mm-hmm. And so what we really liked about Nvidia was that it felt like we just got a lot more resources to do all of that.I think, uh, you [00:08:00] know, NVIDIA's goal is to make things as easy for developers as possible. So there was a really nice like synergy there. I think that, you know, when it comes to like an acquisition, I think the amount that the soul of the products align, I think is gonna be. Is going speak to the success of the acquisition.Yeah. And so it in many ways feels like we're home. This is a really great outcome for us. Like we you know, I love brev.nvidia.com. Like you should, you should use it's, it's theKyle: front page for GPUs.Nader: Yeah. Yeah. If you want GP views,Kyle: you go there, getswyx: it there, and it's like internally is growing very quickly.I, I don't remember You said some stats there.Nader: Yeah, yeah, yeah. It's, uh, I, I wish I had the exact numbers, but like internally, externally, it's been growing really quickly. We've been working with a bunch of partners with a bunch of different customers and ISVs, if you have a solution that you want someone that runs on the GPU and you want people to use it quickly, we can bundle it up, uh, in a launchable and make it a one click run.If you're doing things and you want just like a sandbox or something to run on, right. Like open claw. Huge moment. Super exciting. Our, uh, and we'll talk into it more, but. You know, internally, people wanna run this, and you, we know we have to be really careful from the security implications. Do we let this run on the corporate network?Security's guidance was, Hey, [00:09:00] run this on breath, it's in, you know, it's, it's, it's a vm, it's sitting in the cloud, it's off the corporate network. It's isolated. And so that's been our stance internally and externally about how to even run something like open call while we figure out how to run these things securely.But yeah,swyx: I think there's also like, you almost like we're the right team at the right time when Nvidia is starting to invest a lot more in developer experience or whatever you call it. Yeah. Uh, UX or I don't know what you call it, like software. Like obviously NVIDIA is always invested in software, but like, there's like, this is like a different audience.Yeah. It's aNader: widerKyle: developer base.swyx: Yeah. Right.Nader: Yeah. Yeah. You know, it's funny, it's like, it's not, uh,swyx: so like, what, what is it called internally? What, what is this that people should be aware that is going on there?Nader: Uh, what, like developer experienceswyx: or, yeah, yeah. Is it's called just developer experience or is there like a broader strategy hereNader: in Nvidia?Um, Nvidia always wants to make a good developer experience. The thing is and a lot of the technology is just really complicated. Like, it's not, it's uh, you know, I think, um. The thing that's been really growing or the AI's growing is having a huge moment, not [00:10:00] because like, let's say data scientists in 2018, were quiet then and are much louder now.The pie is com, right? There's a whole bunch of new audiences. My mom's wondering what she's doing. My sister's learned, like taught herself how to code. Like the, um, you know, I, I actually think just generally AI's a big equalizer and you're seeing a more like technologically literate society, I guess.Like everyone's, everyone's learning how to code. Uh, there isn't really an excuse for that. And so building a good UX means that you really understand who your end user is. And when your end user becomes such a wide, uh, variety of people, then you have to almost like reinvent the practice, right? Yeah. You haveKyle: to, and actually build more developer ux, right?Because the, there are tiers of developer base that were added. You know, the, the hackers that are building on top of open claw, right? For example, have never used gpu. They don't know what kuda is. They, they, they just want to run something.Nader: Yeah.Kyle: You need new UX that is not just. Hey, you know, how do you program something in Cuda and run it?And then, and then we built, you know, like when Deep Learning was getting big, we built, we built Torch and, and, but so recently the amount of like [00:11:00] layers that are added to that developer stack has just exploded because AI has become ubiquitous. Everyone's using it in different ways. Yeah. It'sNader: moving fast in every direction.Vertical, horizontal.Vibhu: Yeah. You guys, you even take it down to hardware, like the DGX Spark, you know, it's, it's basically the same system as just throwing it up on big GPU cluster.Nader: Yeah, yeah, yeah. It's amazing. Blackwell.swyx: Yeah. Uh, we saw the preview at the last year's GTC and that was one of the better performing, uh, videos so far, and video coverage so far.Awesome. This will beat it. Um,Nader: that wasswyx: actually, we have fingersNader: crossed. Yeah.DGX Spark and Remote AccessNader: Even when Grace Blackwell or when, um, uh, DGX Spark was first coming out getting to be involved in that from the beginning of the developer experience. And it just comes back to what youswyx: were involved.Nader: Yeah. St. St.swyx: Mars.Nader: Yeah. Yeah. I mean from, it was just like, I, I got an email, we just got thrown into the loop and suddenly yeah, I, it was actually really funny ‘cause I'm still pretty fresh from the acquisition and I'm, I'm getting an email from a bunch of the engineering VPs about like, the new hardware, GPU chip, like we're, or not chip, but just GPU system that we're putting out.And I'm like, okay, cool. Matters. Now involved with this for the ux, I'm like. What am I gonna do [00:12:00] here? So, I remember the first meeting, I was just like kind of quiet as I was hearing engineering VPs talk about what this box could be, what it could do, how we should use it. And I remember, uh, one of the first ideas that people were idea was like, oh, the first thing that it was like, I think a quote was like, the first thing someone's gonna wanna do with this is get two of them and run a Kubernetes cluster on top of them.And I was like, oh, I think I know why I'm here. I was like, the first thing we're doing is easy. SSH into the machine. And then, and you know, just kind of like scoping it down of like, once you can do that every, you, like the person who wants to run a Kubernetes cluster onto Sparks has a higher propensity for pain, then, then you know someone who buys it and wants to run open Claw right now, right?If you can make sure that that's as effortless as possible, then the rest becomes easy. So there's a tool called Nvidia Sync. It just makes the SSH connection really simple. So, you know, if you think about it like. If you have a Mac, uh, or a PC or whatever, if you have a laptop and you buy this GPU and you want to use it, you should be able to use it like it's A-A-G-P-U in the cloud, right?Um, but there's all this friction of like, how do you actually get into that? That's part of [00:13:00] Revs value proposition is just, you know, there's a CLI that wraps SSH and makes it simple. And so our goal is just get you into that machine really easily. And one thing we just launched at CES, it's in, it's still in like early access.We're ironing out some kinks, but it should be ready by GTC. You can register your spark on Brev. And so now if youswyx: like remote managed yeah, local hardware. Single pane of glass. Yeah. Yeah. Because Brev can already manage other clouds anyway, right?Vibhu: Yeah, yeah. And you use the spark on Brev as well, right?Nader: Yeah. But yeah, exactly. So, so you, you, so you, you set it up at home you can run the command on it, and then it gets it's essentially it'll appear in your Brev account, and then you can take your laptop to a Starbucks or to a cafe, and you'll continue to use your, you can continue use your spark just like any other cloud node on Brev.Yeah. Yeah. And it's just like a pre-provisioned centerswyx: in yourNader: home. Yeah, exactly.swyx: Yeah. Yeah.Vibhu: Tiny little data center.Nader: Tiny little, the size ofVibhu: your phone.SOL Culture and Dynamo Setupswyx: One more thing before we move on to Kyle. Just have so many Jensen stories and I just love, love mining Jensen stories. Uh, my favorite so far is SOL. Uh, what is, yeah, what is S-O-L-S-O-LNader: is actually, i, I think [00:14:00] of all the lessons I've learned, that one's definitely my favorite.Kyle: It'll always stick with you.Nader: Yeah. Yeah. I, you know, in your startup, everything's existential, right? Like we've, we've run out of money. We were like, on the risk of, of losing payroll, we've had to contract our team because we l ran outta money. And so like, um, because of that you're really always forcing yourself to I to like understand the root cause of everything.If you get a date, if you get a timeline, you know exactly why that date or timeline is there. You're, you're pushing every boundary and like, you're not just say, you're not just accepting like a, a no. Just because. And so as you start to introduce more layers, as you start to become a much larger organization, SOL is is essentially like what is the physics, right?The speed of light moves at a certain speed. So if flight's moving some slower, then you know something's in the way. So before trying to like layer reality back in of like, why can't this be delivered at some date? Let's just understand the physics. What is the theoretical limit to like, uh, how fast this can go?And then start to tell me why. ‘cause otherwise people will start telling you why something can't be done. But actually I think any great leader's goal is just to create urgency. Yeah. [00:15:00] There's an infiniteKyle: create compelling events, right?Nader: Yeah.Kyle: Yeah. So l is a term video is used to instigate a compelling event.You say this is done. How do we get there? What is the minimum? As much as necessary, as little as possible thing that it takes for us to get exactly here and. It helps you just break through a bunch of noise.swyx: Yeah.Kyle: Instantly.swyx: One thing I'm unclear about is, can only Jensen use the SOL card? Like, oh, no, no, no.Not everyone get the b******t out because obviously it's Jensen, but like, can someone else be like, no, likeKyle: frontline engineers use it.Nader: Yeah. Every, I think it's not so much about like, get the b******t out. It's like, it's like, give me the root understanding, right? Like, if you tell me something takes three weeks, it like, well, what's the first principles?Yeah, the first principles. It's like, what's the, what? Like why is it three weeks? What is the actual yeah. What's the actual limit of why this is gonna take three weeks? If you're gonna, if you, if let's say you wanted to buy a new computer and someone told you it's gonna be here in five days, what's the SOL?Well, like the SOL is like, I could walk into a Best Buy and pick it up for you. Right? So then anything that's like beyond that is, and is that practical? Is that how we're gonna, you know, let's say give everyone in the [00:16:00] company a laptop, like obviously not. So then like that's the SOL and then it's like, okay, well if we have to get more than 10, suddenly there might be some, right?And so now we can kind of piece the reality back.swyx: So, so this is the. Paul Graham do things that don't scale. Yeah. And this is also the, what people would now call behi agency. Yeah.Kyle: It's actually really interesting because there's a, there's a second hardware angle to SOL that like doesn't come up for all the org sol is used like culturally at aswyx: media for everything.I'm also mining for like, I think that can be annoying sometimes. And like someone keeps going IOO you and you're like, guys, like we have to be stable. We have to, we to f*****g plan. Yeah.Kyle: It's an interesting balance.Nader: Yeah. I encounter that with like, actually just with, with Alec, right? ‘cause we, we have a new conference so we need to launch, we have, we have goals of what we wanna launch by, uh, by the conference and like, yeah.At the end of the day, where isswyx: this GTC?Nader: Um, well this is like, so we, I mean we did it for CES, we did for GT CDC before that we're doing it for GTC San Jose. So I mean, like every, you know, we have a new moment. Um, and we want to launch something. Yeah. And we want to do so at SOL and that does mean that some, there's some level of prioritization that needs [00:17:00] to happen.And so it, it is difficult, right? I think, um, you have to be careful with what you're pushing. You know, stability is important and that should be factored into S-O-L-S-O-L isn't just like, build everything and let it break, you know, that, that's part of the conversation. So as you're laying, layering in all the details, one of them might be, Hey, we could build this, but then it's not gonna be stable for X, y, z reasons.And so that was like, one of our conversations for CES was, you know, hey, like we, we can get this into early access registering your spark with brev. But there are a lot of things that we need to do in order to feel really comfortable from a security perspective, right? There's a lot of networking involved before we deliver that to users.So it's like, okay. Let's get this to a point where we can at least let people experiment with it. We had it in a booth, we had it in Jensen's keynote, and then let's go iron out all the networking kinks. And that's not easy. And so, uh, that can come later. And so that was the way that we layered that back in.Yeah. ButKyle: It's not really about saying like, you don't have to do the, the maintenance or operational work. It's more about saying, you know, it's kind of like [00:18:00] highlights how progress is incremental, right? Like, what is the minimum thing that we can get to. And then there's SOL for like every component after that.But there's the SOL to get you, get you to the, the starting line. And that, that's usually how it's asked. Yeah. On the other side, you know, like SOL came out of like hardware at Nvidia. Right. So SOL is like literally if we ran the accelerator or the GPU with like at basically full speed with like no other constraints, like how FAST would be able to make a program go.swyx: Yeah. Yeah. Right.Kyle: Soswyx: in, in training that like, you know, then you work back to like some percentage of like MFU for example.Kyle: Yeah, that's a, that's a great example. So like, there's an, there's an S-O-L-M-F-U, and then there's like, you know, what's practically achievable.swyx: Cool. Should we move on to sort of, uh, Kyle's side?Uh, Kyle, you're coming more from the data science world. And, uh, I, I mean I always, whenever, whenever I meet someone who's done working in tabular stuff, graph neural networks, time series, these are basically when I go to new reps, I go to ICML, I walk the back halls. There's always like a small group of graph people.Yes. Absolute small group of tabular people. [00:19:00] And like, there's no one there. And like, it's very like, you know what I mean? Like, yeah, no, like it's, it's important interesting work if you care about solving the problems that they solve.Kyle: Yeah.swyx: But everyone else is just LMS all the time.Kyle: Yeah. I mean it's like, it's like the black hole, right?Has the event horizon reached this yet in nerves? Um,swyx: but like, you know, those are, those are transformers too. Yeah. And, and those are also like interesting things. Anyway, uh, I just wanted to spend a little bit of time on, on those, that background before we go into Dynamo, uh, proper.Kyle: Yeah, sure. I took a different path to Nvidia than that, or I joined six years ago, seven, if you count, when I was an intern.So I joined Nvidia, like right outta college. And the first thing I jumped into was not what I'd done in, during internship, which was like, you know, like some stuff for autonomous vehicles, like heavyweight object detection. I jumped into like, you know, something, I'm like, recommenders, this is popular. Andswyx: yeah, he did RexiKyle: as well.Yeah, Rexi. Yeah. I mean that, that was the taboo data at the time, right? You have tables of like, audience qualities and item qualities, and you're trying to figure out like which member of [00:20:00] the audience matches which item or, or more practically which item matches which member of the audience. And at the time, really it was like we were trying to enable.Uh, recommender, which had historically been like a little bit of a CP based workflow into something that like, ran really well in GPUs. And it's since been done. Like there are a bunch of libraries for Axis that run on GPUs. Uh, the common models like Deeplearning recommendation model, which came outta meta and the wide and deep model, which was used or was released by Google were very accelerated by GPUs using, you know, the fast HBM on the chips, especially to do, you know, vector lookups.But it was very interesting at the time and super, super relevant because like we were starting to get like. This explosion of feeds and things that required rec recommenders to just actively be on all the time. And sort of transitioned that a little bit towards graph neural networks when I discovered them because I was like, okay, you can actually use graphical neural networks to represent like, relationships between people, items, concepts, and that, that interested me.So I jumped into that at [00:21:00] Nvidia and, and got really involved for like two-ish years.swyx: Yeah. Uh, and something I learned from Brian Zaro Yeah. Is that you can just kind of choose your own path in Nvidia.Kyle: Oh my God. Yeah.swyx: Which is not a normal big Corp thing. Yeah. Like you, you have a lane, you stay in your lane.Nader: I think probably the reason why I enjoy being in a, a big company, the mission is the boss probably from a startup guy. Yeah. The missionswyx: is the boss.Nader: Yeah. Uh, it feels like a big game of pickup basketball. Like, you know, if you play one, if you wanna play basketball, you just go up to the court and you're like, Hey look, we're gonna play this game and we need three.Yeah. And you just like find your three. That's honestly for every new initiative that's what it feels like. Yeah.Vibhu: It also like shows, right? Like Nvidia. Just releasing state-of-the-art stuff in every domain. Yeah. Like, okay, you expect foundation models with Nemo tron voice just randomly parakeet.Call parakeet just comes out another one, uh, voice. TheKyle: video voice team has always been producing.Vibhu: Yeah. There's always just every other domain of paper that comes out, dataset that comes out. It's like, I mean, it also stems back to what Nvidia has to do, right? You have to make chips years before they're actually produced.Right? So you need to know, you need to really [00:22:00] focus. TheKyle: design process starts likeVibhu: exactlyKyle: three to five years before the chip gets to the market.Vibhu: Yeah. I, I'm curious more about what that's like, right? So like, you have specialist teams. Is it just like, you know, people find an interest, you go in, you go deep on whatever, and that kind of feeds back into, you know, okay, we, we expect predictions.Like the internals at Nvidia must be crazy. Right? You know? Yeah. Yeah. You know, you, you must. Not even without selling to people, you have your own predictions of where things are going. Yeah. And they're very based, very grounded. Right?Kyle: Yeah. It, it, it's really interesting. So there's like two things that I think that Amed does, which are quite interesting.Uh, one is like, we really index into passion. There's a big. Sort of organizational top sound push to like ensure that people are working on the things that they're passionate about. So if someone proposes something that's interesting, many times they can just email someone like way up the chain that they would find this relevant and say like, Hey, can I go work on this?Nader: It's actually like I worked at a, a big company for a couple years before, uh, starting on my startup journey and like, it felt very weird if you were to like email out of chain, if that makes [00:23:00] sense. Yeah. The emails at Nvidia are like mosh pitsswyx: shoot,Nader: and it's just like 60 people, just whatever. And like they're, there's this,swyx: they got messy like, reply all you,Nader: oh, it's in, it's insane.It's insane. They justKyle: help. You know, Maxim,Nader: the context. But, but that's actually like, I've actually, so this is a weird thing where I used to be like, why would we send emails? We have Slack. I am the entire, I'm the exact opposite. I feel so bad for anyone who's like messaging me on Slack ‘cause I'm so unresponsive.swyx: Your emailNader: Maxi, email Maxim. I'm email maxing Now email is a different, email is perfect because man, we can't work together. I'm email is great, right? Because important threads get bumped back up, right? Yeah, yeah. Um, and so Slack doesn't do that. So I just have like this casino going off on the right or on the left and like, I don't know which thread was from where or what, but like the threads get And then also just like the subject, so you can have like working threads.I think what's difficult is like when you're small, if you're just not 40,000 people I think Slack will work fine, but there's, I don't know what the inflection point is. There is gonna be a point where that becomes really messy and you'll actually prefer having email. ‘cause you can have working threads.You can cc more than nine people in a thread.Kyle: You can fork stuff.Nader: You can [00:24:00] fork stuff, which is super nice and just like y Yeah. And so, but that is part of where you can propose a plan. You can also just. Start, honestly, momentum's the only authority, right? So like, if you can just start, start to make a little bit of progress and show someone something, and then they can try it.That's, I think what's been, you know, I think the most effective way to push anything for forward. And that's both at Nvidia and I think just generally.Kyle: Yeah, there's, there's the other concept that like is explored a lot at Nvidia, which is this idea of a zero billion dollar business. Like market creation is a big thing at Nvidia.Like,swyx: oh, you want to go and start a zero billion dollar business?Kyle: Jensen says, we are completely happy investing in zero billion dollar markets. We don't care if this creates revenue. It's important for us to know about this market. We think it will be important in the future. It can be zero billion dollars for a while.I'm probably minging as words here for, but like, you know, like, I'll give an example. NVIDIA's been working on autonomous driving for a a long time,swyx: like an Nvidia car.Kyle: No, they, they'veVibhu: used the Mercedes, right? They're around the HQ and I think it finally just got licensed out. Now they're starting to be used quite a [00:25:00] bit.For 10 years you've been seeing Mercedes with Nvidia logos driving.Kyle: If you're in like the South San Santa Clara, it's, it's actually from South. Yeah. So, um. Zero billion dollar markets are, are a thing like, you know, Jensen,swyx: I mean, okay, look, cars are not a zero billion dollar market. But yeah, that's a bad example.Nader: I think, I think he's, he's messaging, uh, zero today, but, or even like internally, right? Like, like it's like, uh, an org doesn't have to ruthlessly find revenue very quickly to justify their existence. Right. Like a lot of the important research, a lot of the important technology being developed that, that's kind ofKyle: where research, research is very ide ideologically free at Nvidia.Yeah. Like they can pursue things that they wereswyx: Were you research officially?Kyle: I was never in research. Officially. I was always in engineering. Yeah. We in, I'm in an org called Deep Warning Algorithms, which is basically just how do we make things that are relevant to deep warning go fast.swyx: That sounds freaking cool.Vibhu: And I think a lot of that is underappreciated, right? Like time series. This week Google put out time. FF paper. Yeah. A new time series, paper res. Uh, Symantec, ID [00:26:00] started applying Transformers LMS to Yes. Rec system. Yes. And when you think the scale of companies deploying these right. Amazon recommendations, Google web search, it's like, it's huge scale andKyle: Yeah.Vibhu: You want fast?Kyle: Yeah. Yeah. Yeah. Actually it's, it, I, there's a fun moment that brought me like full circle. Like, uh, Amazon Ads recently gave a talk where they talked about using Dynamo for generative recommendation, which was like super, like weirdly cathartic for me. I'm like, oh my God. I've, I've supplanted what I was working on.Like, I, you're using LMS now to do what I was doing five years ago.swyx: Yeah. Amazing. And let's go right into Dynamo. Uh, maybe introduce Yeah, sure. To the top down and Yeah.Kyle: I think at this point a lot of people are familiar with the term of inference. Like funnily enough, like I went from, you know, inference being like a really niche topic to being something that's like discussed on like normal people's Twitter feeds.It's,Nader: it's on billboardsKyle: here now. Yeah. Very, very strange. Driving, driving, seeing just an inference ad on 1 0 1 inference at scale is becoming a lot more important. Uh, we have these moments like, you know, open claw where you have these [00:27:00] agents that take lots and lots of tokens, but produce, incredible results.There are many different aspects of test time scaling so that, you know, you can use more inference to generate a better result than if you were to use like a short amount of inference. There's reasoning, there's quiring, there's, adding agency to the model, allowing it to call tools and use skills.Dyno sort came about at Nvidia. Because myself and a couple others were, were sort of talking about the, these concepts that like, you know, you have inference engines like VLMS, shelan, tenor, TLM and they have like one single copy. They, they, they sort of think about like things as like one single copy, like one replica, right?Why Scale Out WinsKyle: Like one version of the model. But when you're actually serving things at scale, you can't just scale up that replica because you end up with like performance problems. There's a scaling limit to scaling up replicas. So you actually have to scale out to use a, maybe some Kubernetes type terminology.We kind of realized that there was like. A lot of potential optimization that we could do in scaling out and building systems for data [00:28:00] center scale inference. So Dynamo is this data center scale inference engine that sits on top of the frameworks like VLM Shilling and 10 T lm and just makes things go faster because you can leverage the economy of scale.The fact that you have KV cash, which we can define a little bit later, uh, in all these machines that is like unique and you wanna figure out like the ways to maximize your cash hits or you want to employ new techniques in inference like disaggregation, which Dynamo had introduced to the world in, in, in March, not introduced, it was a academic talk, but beforehand.But we are, you know, one of the first frameworks to start, supporting it. And we wanna like, sort of combine all these techniques into sort of a modular framework that allows you to. Accelerate your inference at scale.Nader: By the way, Kyle and I became friends on my first date, Nvidia, and I always loved, ‘cause like he always teaches meswyx: new things.Yeah. By the way, this is why I wanted to put two of you together. I was like, yeah, this is, this is gonna beKyle: good. It's very, it's very different, you know, like we've, we, we've, we've talked to each other a bunch [00:29:00] actually, you asked like, why, why can't we scale up?Nader: Yeah.Scale Up Limits ExplainedNader: model, you said model replicas.Kyle: Yeah. So you, so scale up means assigning moreswyx: heavier?Kyle: Yeah, heavier. Like making things heavier. Yeah, adding more GPUs. Adding more CPUs. Scale out is just like having a barrier saying, I'm gonna duplicate my representation of the model or a representation of this microservice or something, and I'm gonna like, replicate it Many times.Handle, load. And the reason that you can't scale, scale up, uh, past some points is like, you know, there, there, there are sort of hardware bounds and algorithmic bounds on, on that type of scaling. So I'll give you a good example that's like very trivial. Let's say you're on an H 100. The Maxim ENV link domain for H 100, for most Ds H one hundreds is heus, right?So if you scaled up past that, you're gonna have to figure out ways to handle the fact that now for the GPUs to communicate, you have to do it over Infin band, which is still very fast, but is not as fast as ENV link.swyx: Is it like one order of magnitude, like hundreds or,Kyle: it's about an order of magnitude?Yeah. Okay. Um, soswyx: not terrible.Kyle: [00:30:00] Yeah. I, I need to, I need to remember the, the data sheet here, like, I think it's like about 500 gigabytes. Uh, a second unidirectional for ENV link, and about 50 gigabytes a second unidirectional for Infin Band. I, it, it depends on the, the generation.swyx: I just wanna set this up for people who are not familiar with these kinds of like layers and the trash speedVibhu: and all that.Of course.From Laptop to Multi NodeVibhu: Also, maybe even just going like a few steps back before that, like most people are very familiar with. You see a, you know, you can use on your laptop, whatever these steel viol, lm you can just run inference there. All, there's all, you can, youcan run it on thatVibhu: laptop. You can run on laptop.Then you get to, okay, uh, models got pretty big, right? JLM five, they doubled the size, so mm-hmm. Uh, what do you do when you have to go from, okay, I can get 128 gigs of memory. I can run it on a spark. Then you have to go multi GPU. Yeah. Okay. Multi GPU, there's some support there. Now, if I'm a company and I don't have like.I'm not hiring the best researchers for this. Right. But I need to go [00:31:00] multi-node, right? I have a lot of servers. Okay, now there's efficiency problems, right? You can have multiple eight H 100 nodes, but, you know, is that as a, like, how do you do that efficiently?Kyle: Yeah. How do you like represent them? How do you choose how to represent the model?Yeah, exactly right. That's a, that's like a hard question. Everyone asks, how do you size oh, I wanna run GLM five, which just came out new model. There have been like four of them in the past week, by the way, like a bunch of new models.swyx: You know why? Right? Deep seek.Kyle: No comment. Oh. Yeah, but Ggl, LM five, right?We, we have this, new model. It's, it's like a large size, and you have to figure out how to both scale up and scale out, right? Because you have to find the right representation that you care about. Everyone does this differently. Let's be very clear. Everyone figures this out in their own path.Nader: I feel like a lot of AI or ML even is like, is like this. I think people think, you know, I, I was, there was some tweet a few months ago that was like, why hasn't fine tuning as a service taken off? You know, that might be me. It might have been you. Yeah. But people want it to be such an easy recipe to follow.But even like if you look at an ML model and specificKyle: to you Yeah,Nader: yeah.Kyle: And the [00:32:00] model,Nader: the situation, and there's just so much tinkering, right? Like when you see a model that has however many experts in the ME model, it's like, why that many experts? I don't, they, you know, they tried a bunch of things and that one seemed to do better.I think when it comes to how you're serving inference, you know, you have a bunch of decisions to make and there you can always argue that you can take something and make it more optimal. But I think it's this internal calibration and appetite for continued calibration.Vibhu: Yeah. And that doesn't mean like, you know, people aren't taking a shot at this, like tinker from thinking machines, you know?Yeah. RL as a service. Yeah, totally. It's, it also gets even harder when you try to do big model training, right? We're not the best at training Moes, uh, when they're pre-trained. Like we saw this with LAMA three, right? They're trained in such a sparse way that meta knows there's gonna be a bunch of inference done on these, right?They'll open source it, but it's very trained for what meta infrastructure wants, right? They wanna, they wanna inference it a lot. Now the question to basically think about is, okay, say you wanna serve a chat application, a coding copilot, right? You're doing a layer of rl, you're serving a model for X amount of people.Is it a chat model, a coding model? Dynamo, you know, back to that,Kyle: it's [00:33:00] like, yeah, sorry. So you we, we sort of like jumped off of, you know, jumped, uh, on that topic. Everyone has like, their own, own journey.Cost Quality Latency TradeoffsKyle: And I, I like to think of it as defined by like, what is the model you need? What is the accuracy you need?Actually I talked to NA about this earlier. There's three axes you care about. What is the quality that you're able to produce? So like, are you accurate enough or can you complete the task with enough, performance, high enough performance. Yeah, yeah. Uh, there's cost. Can you serve the model or serve your workflow?Because it's not just the model anymore, it's the workflow. It's the multi turn with an agent cheaply enough. And then can you serve it fast enough? And we're seeing all three of these, like, play out, like we saw, we saw new models from OpenAI that you know, are faster. You have like these new fast versions of models.You can change the amount of thinking to change the amount of quality, right? Produce more tokens, but at a higher cost in a, in a higher latency. And really like when you start this journey of like trying to figure out how you wanna host a model, you, you, you think about three things. What is the model I need to serve?How many times do I need to call it? What is the input sequence link was [00:34:00] the, what does the workflow look like on top of it? What is the SLA, what is the latency SLA that I need to achieve? Because there's usually some, this is usually like a constant, you, you know, the SLA that you need to hit and then like you try and find the lowest cost version that hits all of these constraints.Usually, you know, you, you start with those things and you say you, you kind of do like a bit of experimentation across some common configurations. You change the tensor parallel size, which is a form of parallelismVibhu: I take, it goes even deeper first. Gotta think what model.Kyle: Yes, course,ofKyle: course. It's like, it's like a multi-step design process because as you said, you can, you can choose a smaller model and then do more test time scaling and it'll equate the quality of a larger model because you're doing the test time scaling or you're adding a harness or something.So yes, it, it goes way deeper than that. But from the performance perspective, like once you get to the model you need, you need to host, you look at that and you say, Hey. I have this model, I need to serve it at the speed. What is the right configuration for that?Nader: You guys see the recent, uh, there was a paper I just saw like a few days ago that, uh, if you run [00:35:00] the same prompt twice, you're getting like double Just try itagain.Nader: Yeah, exactly.Vibhu: And you get a lot. Yeah. But the, the key thing there is you give the context of the failed try, right? Yeah. So it takes a shot. And this has been like, you know, basic guidance for quite a while. Just try again. ‘cause you know, trying, just try again. Did you try again? All adviceNader: in life.Vibhu: Just, it's a paper from Google, if I'm not mistaken, right?Yeah,Vibhu: yeah. I think it, it's like a seven bas little short paper. Yeah. Yeah. The title's very cute. And it's just like, yeah, just try again. Give it ask context,Kyle: multi-shot. You just like, say like, hey, like, you know, like take, take a little bit more, take a little bit more information, try and fail. Fail.Vibhu: And that basic concept has gone pretty deep.There's like, um, self distillation, rl where you, you do self distillation, you do rl and you have past failure and you know, that gives some signal so people take, try it again. Not strong enough.swyx: Uh, for, for listeners, uh, who listen to here, uh, vivo actually, and I, and we run a second YouTube channel for our paper club where, oh, that's awesome.Vivo just covered this. Yeah. Awesome. Self desolation and all that's, that's why he, to speed [00:36:00] on it.Nader: I'll to check it out.swyx: Yeah. It, it's just a good practice, like everyone needs, like a paper club where like you just read papers together and the social pressure just kind of forces you to just,Nader: we, we,there'sNader: like a big inference.Kyle: ReadingNader: group at a video. I feel so bad every time. I I, he put it on like, on our, he shared it.swyx: One, one ofNader: your guys,swyx: uh, is, is big in that, I forget es han Yeah, yeah,Kyle: es Han's on my team. Actually. Funny. There's a, there's a, there's a employee transfer between us. Han worked for Nater at Brev, and now he, he's on my team.He wasNader: our head of ai. And then, yeah, once we got in, andswyx: because I'm always looking for like, okay, can, can I start at another podcast that only does that thing? Yeah. And, uh, Esan was like, I was trying to like nudge Esan into like, is there something here? I mean, I don't think there's, there's new infant techniques every day.So it's like, it's likeKyle: you would, you would actually be surprised, um, the amount of blog posts you see. And ifswyx: there's a period where it was like, Medusa hydra, what Eagle, like, youKyle: know, now we have new forms of decode, uh, we have new forms of specula, of decoding or new,swyx: what,Kyle: what are youVibhu: excited? And it's exciting when you guys put out something like Tron.‘cause I remember the paper on this Tron three, [00:37:00] uh, the amount of like post train, the on tokens that the GPU rich can just train on. And it, it was a hybrid state space model, right? Yeah.Kyle: It's co-designed for the hardware.Vibhu: Yeah, go design for the hardware. And one of the things was always, you know, the state space models don't scale as well when you do a conversion or whatever the performance.And you guys are like, no, just keep draining. And Nitron shows a lot of that. Yeah.Nader: Also, something cool about Nitron it was released in layers, if you will, very similar to Dynamo. It's, it's, it's essentially it was released as you can, the pre-training, post-training data sets are released. Yeah. The recipes on how to do it are released.The model itself is released. It's full model. You just benefit from us turning on the GPUs. But there are companies like, uh, ServiceNow took the dataset and they trained their own model and we were super excited and like, you know, celebrated that work.ZoomVibhu: different. Zoom is, zoom is CGI, I think, uh, you know, also just to add like a lot of models don't put out based models and if there's that, why is fine tuning not taken off?You know, you can do your own training. Yeah,Kyle: sure.Vibhu: You guys put out based model, I think you put out everything.Nader: I believe I know [00:38:00]swyx: about base. BasicallyVibhu: without baseswyx: basic can be cancelable.Vibhu: Yeah. Base can be cancelable.swyx: Yeah.Vibhu: Safety training.swyx: Did we get a full picture of dymo? I, I don't know if we, what,Nader: what I'd love is you, you mentioned the three axes like break it down of like, you know, what's prefilled decode and like what are the optimizations that we can get with Dynamo?Kyle: Yeah. That, that's, that's, that's a great point. So to summarize on that three axis problem, right, there are three things that determine whether or not something can be done with inference, cost, quality, latency, right? Dynamo is supposed to be there to provide you like the runtime that allows you to pull levers to, you know, mix it up and move around the parade of frontier or the preto surface that determines is this actually possible with inference And AI todayNader: gives you the knobs.Kyle: Yeah, exactly. It gives you the knobs.Disaggregation Prefill vs DecodeKyle: Uh, and one thing that like we, we use a lot in contemporary inference and is, you know, starting to like pick up from, you know, in, in general knowledge is this co concept of disaggregation. So historically. Models would be hosted with a single inference engine. And that inference engine [00:39:00] would ping pong between two phases.There's prefill where you're reading the sequence generating KV cache, which is basically just a set of vectors that represent the sequence. And then using that KV cache to generate new tokens, which is called Decode. And some brilliant researchers across multiple different papers essentially made the realization that if you separate these two phases, you actually gain some benefits.Those benefits are basically a you don't have to worry about step synchronous scheduling. So the way that an inference engine works is you do one step and then you finish it, and then you schedule, you start scheduling the next step there. It's not like fully asynchronous. And the problem with that is you would have, uh, essentially pre-fill and decode are, are actually very different in terms of both their resource requirements and their sometimes their runtime.So you would have like prefill that would like block decode steps because you, you'd still be pre-filing and you couldn't schedule because you know the step has to end. So you remove that scheduling issue and then you also allow you, or you yourself, to like [00:40:00] split the work into two different ki types of pools.So pre-fill typically, and, and this changes as, as model architecture changes. Pre-fill is, right now, compute bound most of the time with the sequence is sufficiently long. It's compute bound. On the decode side because you're doing a full Passover, all the weights and the entire sequence, every time you do a decode step and you're, you don't have the quadratic computation of KV cache, it's usually memory bound because you're retrieving a linear amount of memory and you're doing a linear amount of compute as opposed to prefill where you retrieve a linear amount of memory and then use a quadratic.You know,Nader: it's funny, someone exo Labs did a really cool demo where for the DGX Spark, which has a lot more compute, you can do the pre the compute hungry prefill on a DG X spark and then do the decode on a, on a Mac. Yeah. And soVibhu: that's faster.Nader: Yeah. Yeah.Kyle: So you could, you can do that. You can do machine strat stratification.Nader: Yeah.Kyle: And like with our future generation generations of hardware, we actually announced, like with Reuben, this [00:41:00] new accelerator that is prefilled specific. It's called Reuben, CPX. SoKubernetes Scaling with GroveNader: I have a question when you do the scale out. Yeah. Is scaling out easier with Dynamo? Because when you need a new node, you can dedicate it to either the Prefill or, uh, decode.Kyle: Yeah. So Dynamo actually has like a, a Kubernetes component in it called Grove that allows you to, to do this like crazy scaling specialization. It has like this hot, it's a representation that, I don't wanna go too deep into Kubernetes here, but there was a previous way that you would like launch multi-node work.Uh, it's called Leader Worker Set. It's in the Kubernetes standard, and Leader worker set is great. It served a lot of people super well for a long period of time. But one of the things that it's struggles with is representing a set of cases where you have a multi-node replica that has a pair, right?You know, prefill and decode, or it's not paired, but it has like a second stage that has a ratio that changes over time. And prefill and decode are like two different things as your workload changes, right? The amount of prefill you'll need to do may change. [00:42:00] The amount of decode that you, you'll need to do might change, right?Like, let's say you start getting like insanely long queries, right? That probably means that your prefill scales like harder because you're hitting these, this quadratic scaling growth.swyx: Yeah.And then for listeners, like prefill will be long input. Decode would be long output, for example, right?Kyle: Yeah. So like decode, decode scale. I mean, decode is funny because the amount of tokens that you produce scales with the output length, but the amount of work that you do per step scales with the amount of tokens in the context.swyx: Yes.Kyle: So both scales with the input and the output.swyx: That's true.Kyle: But on the pre-fold view code side, like if.Suddenly, like the amount of work you're doing on the decode side stays about the same or like scales a little bit, and then the prefilled side like jumps up a lot. You actually don't want that ratio to be the same. You want it to change over time. So Dynamo has a set of components that A, tell you how to scale.It tells you how many prefilled workers and decoded workers you, it thinks you should have, and also provides a scheduling API for Kubernetes that allows you to actually represent and affect this scheduling on, on, on your actual [00:43:00] hardware, on your compute infrastructure.Nader: Not gonna lie. I feel a little embarrassed for being proud of my SVG function earlier.swyx: No, itNader: wasreallyKyle: cute. I, Iswyx: likeNader: it's all,swyx: it's all engineering. It's all engineering. Um, that's where I'mKyle: technical.swyx: One thing I'm, I'm kind of just curious about with all with you see at a systems level, everything going on here. Mm-hmm. And we, you know, we're scaling it up in, in multi, in distributed systems.Context Length and Co Designswyx: Um, I think one thing that's like kind of, of the moment right now is people are asking, is there any SOL sort of upper bounds. In terms of like, let's call, just call it context length for one for of a better word, but you can break it down however you like.Nader: Yeah.swyx: I just think like, well, yeah, I mean, like clearly you can engage in hybrid architectures and throw in some state space models in there.All, all you want, but it looks, still looks very attention heavy.Kyle: Yes. Uh, yeah. Long context is attention heavy. I mean, we have these hybrid models, um,swyx: to take and most, most models like cap out at a million contexts and that's it. Yeah. Like for the last two years has been it.Kyle: Yeah. The model hardware context co-design thing that we're seeing these days is actually super [00:44:00] interesting.It's like my, my passion, like my secret side passion. We see models like Kimmy or G-P-T-O-S-S. I'm use these because I, I know specific things about these models. So Kimmy two comes out, right? And it's an interesting model. It's like, like a deep seek style architecture is MLA. It's basically deep seek, scaled like a little bit differently, um, and obviously trained differently as well.But they, they talked about, why they made the design choices for context. Kimmy has more experts, but fewer attention heads, and I believe a slightly smaller attention, uh, like dimension. But I need to remember, I need to check that. Uh, it doesn't matter. But they discussed this actually at length in a blog post on ji, which is like our pu which is like credit puswyx: Yeah.Kyle: Um, in, in China. Chinese red.swyx: Yeah.Kyle: It's, yeah. So it, it's, it's actually an incredible blog post. Uh, like all the mls people in, in, in that, I've seen that on GPU are like very brilliant, but they, they talk about like the creators of Kimi K two [00:45:00] actually like, talked about it on, on, on there in the blog post.And they say, we, we actually did an experiment, right? Attention scales with the number of heads, obviously. Like if you have 64 heads versus 32 heads, you do half the work of attention. You still scale quadratic, but you do half the work. And they made a, a very specific like. Sort of barter in their system, in their architecture, they basically said, Hey, what if we gave it more experts, so we're gonna use more memory capacity.But we keep the amount of activated experts the same. We increase the expert sparsity, so we have fewer experts act. The ratio to of experts activated to number of experts is smaller, and we decrease the number of attention heads.Vibhu: And kind of for context, what the, what we had been seeing was you make models sparser instead.So no one was really touching heads. You're just having, uh,Kyle: well, they, they did, they implicitly made it sparser.Vibhu: Yeah, yeah. For, for Kimmy. They did,Kyle: yes.Vibhu: They also made it sparser. But basically what we were seeing was people were at the level of, okay, there's a sparsity ratio. You want more total parameters, less active, and that's sparsity.[00:46:00]But what you see from papers, like, the labs like moonshot deep seek, they go to the level of, okay, outside of just number of experts, you can also change how many attention heads and less attention layers. More attention. Layers. Layers, yeah. Yes, yes. So, and that's all basically coming back to, just tied together is like hardware model, co-design, which isKyle: hardware model, co model, context, co-design.Vibhu: Yeah.Kyle: Right. Like if you were training a, a model that was like. Really, really short context, uh, or like really is good at super short context tasks. You may like design it in a way such that like you don't care about attention scaling because it hasn't hit that, like the turning point where like the quadratic curve takes over.Nader: How do you consider attention or context as a separate part of the co-design? Like I would imagine hardware or just how I would've thought of it is like hardware model. Co-design would be hardware model context co-designKyle: because the harness and the context that is produced by the harness is a part of the model.Once it's trained in,Vibhu: like even though towards the end you'll do long context, you're not changing architecture through I see. Training. Yeah.Kyle: I mean you can try.swyx: You're saying [00:47:00] everyone's training the harness into the model.Kyle: I would say to some degree, orswyx: there's co-design for harness. I know there's a small amount, but I feel like not everyone has like gone full send on this.Kyle: I think, I think I think it's important to internalize the harness that you think the model will be running. Running into the model.swyx: Yeah. Interesting. Okay. Bash is like the universal harness,Kyle: right? Like I'll, I'll give. An example here, right? I mean, or just like a, like a, it's easy proof, right? If you can train against a harness and you're using that harness for everything, wouldn't you just train with the harness to ensure that you get the best possible quality out of,swyx: Well, the, uh, I, I can provide a counter argument.Yeah, sure. Which is what you wanna provide a generally useful model for other people to plug into their harnesses, right? So if youKyle: Yeah. Harnesses can be open, open source, right?swyx: Yeah. So I mean, that's, that's effectively what's happening with Codex.Kyle: Yeah.swyx: And, but like you may want like a different search tool and then you may have to name it differently or,Nader: I don't know how much people have pushed on this, but can you.Train a model, would it be, have you have people compared training a model for the for the harness versus [00:48:00] like post training forswyx: I think it's the same thing. It's the same thing. It's okay. Just extra post training. INader: see.swyx: And so, I mean, cognition does this course, it does this where you, you just have to like, if your tool is slightly different, um, either force your tool to be like the tool that they train for.Hmm. Or undo their training for their tool and then Oh, that's re retrain. Yeah. It's, it's really annoying and like,Kyle: I would hope that eventually we hit like a certain level of generality with respect to training newswyx: tools. This is not a GI like, it's, this is a really stupid like. Learn my tool b***h.Like, I don't know if, I don't know if I can say that, but like, you know, um, I think what my point kind of is, is that there's, like, I look at slopes of the scaling laws and like, this slope is not working, man. We, we are at a million token con

Discord delays their age-gating rollout but legislators are pushing for operating systems including Linux to verify ages, LLM licence laundering might mean the end of copyleft, and how and why you might want to detect Meta’s spy camera glasses. News Getting Global Age Assurance Right: What We Got Wrong and What’s Changing US state laws push age checks into the operating system California introduces age verification law for all operating systems, including Linux and SteamOS — user age verified during OS account setup I have actually read the text of California law CA AB1043 and, honestly, I don’t hate it Do you really think that circumventing these things will always be a simple firmware mod or hardware hack? Relicensing with AI-assisted rewrite – Tuan-Anh Tran Chardet dispute shows how AI will kill software licensing, argues Bruce Perens No right to relicense this project Hide from Meta’s spyglasses with this new Android app Dear Meta Smart Glasses Wearers: You’re Being Watched, Too Automox Turnkey Results Endpoint management tailored to your specific environment. Know the plan. Trust the result. Learn more at www.automox.com Support us on patreon and get an ad-free RSS feed with early episodes sometimes See our contact page for ways to get in touch. RSS: Subscribe to the RSS feeds here

In this conversation, Paul Blankley and Ryan Janssen, founders of Zenlytic, drop in to discuss the massive shift in how we build software and handle data. We trace their journey from studying early NLP and Transformers at Harvard right when the BERT paper dropped, to building a company that relies on cutting-edge LLMs. As far as I know, they're the first to use LLM's for analytics.We dive deep into the reality of the agentic era: engineers are no longer writing the bulk of the code; they are managing agents, verifying outputs, and maintaining ridiculously high standards. We also explore why the industry needs to embrace "net negative scaffolding" as models get smarter, and why having good "taste" might be the ultimate human moat left in tech.Bonus: To prove that software development is changing faster than ever, we literally "vibe coded" a brand-new CRM called "Slop Force" in 20 minutes during this episode. Zenlytic: https://www.zenlytic.com/

Stripe Solves AI Billing, Nvidia's $30B OpenAI Exit, GPT 5.4 Launches with Computer Use, and OpenAI's Safety ReckoningThis week on AI News in 5 by The AI Report, Liam Lawson breaks down four major stories reshaping the AI industry. From Stripe's new billing infrastructure for AI companies to Nvidia's $30 billion investment in OpenAI that may be its last, GPT 5.4 beating 83% of industry professionals, and OpenAI facing a safety crisis after failing to alert law enforcement about a dangerous user.These stories signal a shift in how AI companies monetize products, how the biggest AI labs will fund themselves through public markets, and what safety obligations come with deploying AI at scale. Whether you are building AI products, investing in the space, or deploying enterprise AI, this episode covers the developments you need to know.Key Topics CoveredStripe's new AI billing feature that passes through LLM token costs to customers with automatic markupHow Stripe's tool integrates with third-party gateways like Vercel and OpenRouterNvidia's $30 billion investment in OpenAI as part of the $110 billion funding roundWhy Jensen Huang says the private mega-deal era for AI labs is endingOpenAI's $730 billion valuation and the path to IPO alongside AnthropicGPT 5.4's native computer use capabilities and 1 million token context windowGPT 5.4 benchmark results showing 83% outperformance versus industry professionals33% reduction in factual errors and 47% token savings in tool-heavy workflowsOpenAI's safety crisis after flagging a dangerous user but never contacting law enforcementSam Altman's pledge to overhaul safety protocols including a direct contact line for Canadian policeEpisode Timestamps00:00 - Introduction to AI News in 501:08 - Stripe solves AI's biggest billing problem02:12 - How 30% automated markup works for agentic workflows02:40 - Why unpredictable token costs threaten AI margins03:17 - Stripe launches its own multi-model gateway03:49 - Nvidia's $30 billion OpenAI investment may be its last04:32 - OpenAI and Anthropic gear up for IPOs04:57 - Inside OpenAI's $110 billion funding round and $730 billion valuation05:57 - GPT 5.4 launches with native computer use06:54 - GPT 5.4 benchmarks crush 83% of industry professionals08:55 - OpenAI flagged a dangerous user but never called police09:46 - Sam Altman pledges safety protocol overhaul10:34 - When does a safety flag become a legal obligationResources MentionedStripe AI billing and cost pass-through featureVercel and OpenRouter third-party gateway integrationsNvidia Vera Rubin inference and training systemsOpenAI GPT 5.4 with native computer useChatGPT, Codex, and OpenAI APIChatGPT for Excel add-onMorgan Stanley conference (Jensen Huang keynote)Partner LinksBook Enterprise Training — https://www.upscaile.com/Subscribe to our free newsletter — https://www.theaireport.ai/subscribe-theaireport-youtube#AINews #GPT5 #OpenAI #Nvidia #Stripe #AIBilling #JensenHuang #SamAltman #EnterpriseAI #AISafety #AIAgents #ComputerUse #LLM #AIInfrastructure #TokenCosts

Your prospects stopped Googling. They're asking ChatGPT which CRM to buy, asking Perplexity which marketing automation platform solves their problem, asking Gemini to compare your category - and AI is recommending your competitors, not you. This isn't a "future of search" problem. This is a pipeline problem. Right now. Kevin White, Head of Marketing at Scrunch (and formerly leading growth at Twilio Segment, Retool, and Common Room), saw this crisis coming and built the measurement infrastructure to fix it. His clients are seeing 100% to 260% visibility gains in AI search in just 60 days - and tracking customers who convert specifically because they appeared in ChatGPT answers. In this episode, Kevin deconstructs the "Citation Economy" - why chasing Reddit mentions is a trap, which citation sources actually drive AI recommendations in your vertical, and the exact workflow his team uses to go from "invisible in AI" to "recommended by LLMs" in under two months. You'll learn the diagnostic every Revenue Leader should run Monday morning to find out if your brand exists when prospects ask AI for solutions. You'll discover why the "long, long tail" of AI search (20+ word prompts vs. 5-7 word Google queries) creates greenfield opportunities for brands willing to architect intent-driven content. And you'll hear how Scrunch's Agent Experience Platform delivers AI-optimized content to LLM crawlers without breaking the human website experience. This is not theory. Kevin walks through Scrunch's own dogfooding - how they use their product to track competitive intelligence, diagnose content gaps, and systematically close citation opportunities. He shares the specific technical optimizations that actually move the needle (spoiler: it's not just adding FAQs). And he reveals how agentic workflows - using Claude Code to build marketing apps, automate spreadsheet enrichment, and analyze customer calls - are reclaiming hours his team used to lose to manual synthesis. If your marketing team thinks they're winning because they rank on Page 1 of Google while your pipeline bleeds invisible in ChatGPT, this episode is your wake-up call. Kevin delivers the audit, the workflow, and the hard truth: you can't optimize what you don't measure. Run the diagnostic. Close the gaps. Win the prompts that matter. Show Notes & Transcript: https://theaihat.com/your-prospects-arent-googling-anymore-scrunchs-kevin-white-on-the-ai-visibility-crisis/ CHAPTERS 00:00 LLM SEO Lore 00:36 Podcast Intro Theme 01:45 AI Hat Welcome 02:10 Why Google Rankings Fail 02:53 Kevin White Joins 04:14 Wake Up Call Moment 07:15 Measuring AI Visibility 10:34 Citation Myths Debunked 13:57 Long Long Tail Search 16:45 Three Citation Buckets 19:07 Workflow To 260% Gains 24:15 Shadow AI Ad Break 25:25 Making Sites AI Readable 29:57 Dogfooding Scrunch Tactics 32:19 Top Priorities This Quarter 35:33 One Prompt Audit Trick 37:16 Where To Connect 38:40 Final Wrap Up Kevin White on LinkedIn: https://www.linkedin.com/in/kevbosaurus/Scrunch: https://scrunch.com/ Learn more about your ad choices. Visit megaphone.fm/adchoices

After experiencing Planet Nix and SCaLE, we come back convinced the next phase of Linux is already taking shape.Sponsored By:Jupiter Party Annual Membership: Put your support on automatic with our annual plan, and get one month of membership for free! Managed Nebula: Meet Managed Nebula from Defined Networking. A decentralized VPN built on the open-source Nebula platform that we love. Support LINUX UnpluggedLinks:

You're invited into a legacy family audio business that refused to accept “good enough” on feedback control and instead chased the impossible: a truly zero‑latency, AI‑driven way to push your PA louder without squeals. You follow Devin Sheets from growing up on sound gigs to roaming European stages, then back home to build De‑Feedback plugin for working musicians, a live sound feedback plugin and on‑the‑fly impulse‑response generator that listens like a seasoned engineer: separating human voice, room reverb, background noise, and feedback in real time so you can grab at least 6 dB more gain before things start to howl. Along the way you see how NAMM sparked the idea, how inverse impulse responses and probability math beat old EQ and gate tricks, and how “homebrew AI” meant sneaking into every empty church at 3 a.m. just to teach the model what real rooms actually sound like. You also learn how to think like a modern working musician: using social media to find the right AI programmers across the world, leaning on LLMs to translate, collaborate, and even rate contractor work so you can move faster without losing control. You come away knowing you can drop a dedicated De‑Feedback box or plugin into almost any rig, from churches to touring consoles to tiny clubs, take it with you even when someone else is behind the board, and quietly stack the deck in your favor. In the end, it's a roadmap for how you run your own gigs and career: stay curious, embrace new tools, protect your sound, and Always Be Performing. 00:00:00 Gig Gab 524 – Monday, March 9th, 2026 March 9th: National Meatball Day Guest co-host: Devin Sheets from Alpha Labs 00:02:12 Let's Grow this Legacy Family Business Grew up doing sound Also a musician Lived in Europe Then came back and said, “let's grow this family business!” 00:03:44 We haven't “just solved” this feedback problem Went to NAMM for the first time, and was inspired There are automated EQ-based or gate-based systems PSE plugin from Waves 5045 for feedback 00:04:57 Why isn't there a “balanced audio”-type solution for Feedback Balanced Audio fixes hums and it just works. 00:08:24 NAMM is a great inspiration…and it inspired Devin and his team to seek a feedback plugin solution People get entrenched Inverse Impulse Response methodology 00:12:35 Training the AI to listen for three things: human voice, reverb, and feedback Created a de-reverb algorithm and went beyond that A probability calculation does the math 00:16:05 Truly zero latency for the plugin Workflow latency remains 00:19:32 I don't have any coding or AI background, but I have a gut feeling AI will fix this feedback problem Others: It's harder than you think Devin: I knew that it needed to happen 00:20:58 Finding an AI programmer who was interested in doing Experimented with some programmers, failed, learned some things! 00:21:09 Social Media to the rescue! Late 2023: Devin found a group of AI programmers who would be interested Sending large amounts of money to China…it's a risk! 00:26:30 At 3am, a text message: I think I've done it. Devin immediately started testing it himself “It seemed to work.” 00:27:17 Installing De-Feedback in Churches Sponsors 00:30:57 SPONSOR: Claude.ai – Ready to tackle bigger problems? Sign up for Claude today, which includes access to Claude Cowork, too, when you visit Claude.ai/giggab 00:32:43 SPONSOR: Squarespace. Check out https://www.squarespace.com/GIGGAB to save 10% off your first purchase of a website or domain using code GIGGAB. 00:34:20 What is an impulse response? Impulse Response: An audio picture of how the room sounds Popping balloons in a room/environment and recording the sound is a common approach for creating impulse responses 00:38:33 De-Feedback is an on-the-fly IR generator …and analyzer that's trained on the human voice, room reverb, background noise…and feedback 00:41:55 Finding the right programmers was the key …in addition to actually having the idea and the bullheaded persistence to make it happen. 00:44:46 Mind-melding was necessary And LLMs helped with translation! 00:48:39 Using AI to make it possible to collaborate with other humans 00:50:03 Using an LLM to rate the work of your contractors and employees 00:51:54 How do we get De-Feedback into the hands of working musicians US$499 for the De-Feedback plugin VST3 or AU plugin A higher-end Windows laptop can likely run it on its own Apple's Core Audio tech makes it difficult, but they're working on it. De-Feedback also sells a perfectly-tuned headless computer to do this Alpha Labs tried tons of interfaces that the Focusrite Scarlett keeps glitches out of the mix Waves SuperRack LiveBox 01:01:37 Where do we expand? Allen & Heath mixers? Midas/Behringer mixers? Paul Falcone, mixing Mariah Carey, wanted to use it! Robert Scovill talking Rock Hall on Gig Gab 01:05:18 Homebrew AI! Training EVERY room he could find “Can you let me into your empty church at 3am?” – To record IR to then train the data set for De-Feeback 01:07:25 Creating your own AI model 01:08:13 What's the future look like? Acquisition? Demands for security? – Planning for it all 01:09:26 You can get this and bring it with you to gigs where someone else is doing sound De-Feedback Option 1 Allen & Heath Qu-5's Feedback Eliminator De-Feedback gets at least as 6dB more gain before feedback 01:17:46 Gig Gab 524 Outtro Follow Devin Sheets And Alpha Labs Facebook and Instagram YouTube for Alpha Labs Contact Gig Gab! @GigGabPodcast on Instagram feedback@giggabpodcast.com Sign Up for the Gig Gab Mailing List The post De-Feedback Plugin for Working Musicians: More Gain, Less Feedback – Gig Gab 524 with Devin Sheets appeared first on Gig Gab.

Eoin Clancy (VP of Growth at AirOps), Connor Beaulieu (Senior SEO Manager at LegalZoom), and Adina Timar (Head of AEO at Weflow) join this live session to talk about how to create high-quality content with AI. Connor walks through a workflow his team built at LegalZoom to automatically source expert quotes.. Adina shows how she rebuilds competitor pages from scratch using sales calls, LLM data, and live competitor analysis. And Eoin shares the research behind why content quality is now the single biggest lever in AI search. If you want to see what content engineering actually looks like in practice, this one is for you. Join 50,0000 people who get Dave's Newsletter here: https://www.exitfive.com/newsletterLearn more about Exit Five's private marketing community: https://www.exitfive.com/***Brought to you by:AirOps - The content engineering platform that helps marketers create and maintain high-quality, on-brand content that wins AI search. Go to airops.com/exitfive to start creating content that reflects your expertise, stays true to your brand, and is engineered for performance across human and AI discovery.Customer.io - An AI powered customer engagement platform that help marketers turn first-party data into engaging customer experiences across email, SMS, and push. Learn more at customer.io/exitfive. Convertr - The enterprise lead data management platform that sits between your lead sources and your CRM, automatically validating, enriching, and standardizing every lead before it touches your systems. Check them out at convertr.io/exitfive.Compound Growth Marketing - A full-funnel demand generation agency that helps high-growth cybersecurity, DevOps, and enterprise software companies drive more pipeline through AI SEO, paid media, and go-to-market engineering. Visit compoundgrowthmarketing.com and tell them Dave sent you.***Thanks to my friends at hatch.fm for producing this episode and handling all of the Exit Five podcast production.They give you unlimited podcast editing and strategy for your B2B podcast.Get unlimited podcast editing and on-demand strategy for one low monthly cost. Just upload your episode, and they take care of the rest.Visit hatch.fm to learn more

256 | Andre Alpar ist die deutsche SEO-Legende und Rocket Internet Veteran.Partner dieser Folge:HOLVIFinanzen für kleine Unternehmen: Von Chaos zu Klarheit mit HOLVI - Das kostenlos Holvi Flex Konto ist perfekt für Solopreneure, Freelancer und Unternehmen, die wachsen wollen. ⁠www.holvi.com⁠ClockodoDas Time-Tracking-Tool unserer Wahl. ⁠⁠https://www.clockodo.com/⁠optimisten⁠ Gutschein-Code: optimisten25 für 25% Rabatt.Mach das 1-minütige Quiz und finde eine Geschäftsidee, die zu dir passt: digitaleoptimisten.de/quiz.Learnings**Generative Engines unterscheiden**Im Gespräch wird zwischen dem reinen LLM, einem Chatbot und Overviews unterschieden; diese drei Typen arbeiten unterschiedlich und nutzen unterschiedliche Informationsquellen. Endnutzer erleben dadurch verschiedene Erfahrungen beim Suchen und Antworten. Wer Strategien für AI-gestützte Sichtbarkeit plant, sollte diese Unterschiede kennen, um passende Tools und Vorgehen auszuwählen.**Das Playbook Geo**Im Gespräch wird ein Framework aus Strategie, Technik, Content und Offpage vorgestellt, um in ChatGPT und Co. sichtbar zu werden. Zuerst Strategie klären, dann Technik sicherstellen, Content vorbereiten, Offpage-Signale aufbauen. **Offpage ist entscheidender Hebel**Offpage-Erwähnungen in reputationsstarken Publikationen bleiben relevant. Das Matcha-Tea-Experiment demonstriert, dass Offsite-Aktionen die AI-Sichtbarkeit beeinflussen können. Für Marken bedeutet das: Reputation und Nennung außerhalb der eigenen Seite weiter strategisch pflegen.**AI-Traffic konvertiert besser**AI-Traffic konvertiert 4x besser als organischer Google-Traffic. Gleichzeitig nutzt AI-Systeme den Funnel differenziert (stärkere Nähe zu Entscheidungen). Unternehmen sollten daher AI-Traffic gezielt nutzen, aber Langzeitwirkungen und Abhängigkeiten bedenki; die Qualität der Interaktion ist wichtiger als reiner Traffic.KeywordsGenerative Engine OptimizationAI-gestützte SucheLLM-basierte SucheOffpage-OptimierungContent-Strategie AIwie funktioniert Generative Engine OptimizationUnterschiede zwischen LLM-basierter Suche und konventioneller SucheAuswirkungen von AI Overviews auf CTRJevens Paradox ErklärungMatcha Theos Experiment AI-SEOKnowledge Graph von GoogleMarkennennung als Offsite-Signal in AI-SystemenVertrauen in AI-AntwortenOffpage Signale in AI-Umgebungen

We trace the fast shift from link-based search to AI-generated answers and show how that change reshapes content, measurement, and strategy. Connor Kimball of Cairrot shares data on LLM traffic surges, concrete AEO tactics, and how unified analytics reveals brand lift beyond referral clicks.• AI search replacing link lists with answers• AEO as the method, LLM visibility as the metric• Informational traffic down, transactional content up• Comparison and battle-card pages driving citations• GA4 and GSC integration for unified insights• Measuring halo effects across direct and organic• E‑E‑A‑T signals across site and profiles• Partner ecosystems and regulated-industry expertise• From LLM visibility to multimodal AI visibility• Agents and automated reporting speeding decisionsGuest Contact Information: Website: connorkimball.comLinkedIn: linkedin.com/in/connor-kimballInstagram: instagram.com/connorkimballFacebook: facebook.com/connor.kimballMore from EWR and Matthew:Leave us a review wherever you listen: Spotify, Apple Podcasts, or Amazon PodcastFree SEO Consultation: www.ewrdigital.com/discovery-callWith over 5 million downloads, The Best SEO Podcast has been the go-to show for digital marketers, business owners, and entrepreneurs wanting real-world strategies to grow online. Now, host Matthew Bertram — creator of the LLM Visibility Stack™, and Lead Strategist at EWR Digital — takes the conversation beyond traditional SEO into the AI era of discoverability. Each week, Matthew dives into the tactics, frameworks, and insights that matter most in a world where search engines, large language models, and answer engines are reshaping how people find, trust, and choose businesses. From SEO and AI-driven marketing to executive-level growth strategy, you'll hear expert interviews, deep-dive discussions, and actionable strategies to help you stay ahead of the curve. Find more episodes here: youtube.com/@BestSEOPodcastbestseopodcast.combestseopodcast.buzzsprout.comFollow us on:Facebook: @bestseopodcastInstagram: @thebestseopodcastTiktok: @bestseopodcastLinkedIn: @bestseopodcastConnect With Matthew Bertram: Website: www.matthewbertram.comInstagram: @matt_bertram_liveLinkedIn: @mattbertramlivePowered by: ewrdigital.comSupport the show

Would you like access to our advanced agency training for FREE? https://www.agencymastery360.com/training AI is either the end of agencies… or the biggest opportunity we've had since the internet. Most agree it's the second one. Agencies that are winning right now are combining SEO, GEO, AEO, and LLM optimization so they show up everywhere decisions are being made. They're using AI to increase leverage, not replace thinking. And they're restructuring their teams around strategy, insight, and proprietary data instead of repetitive task work. Today's featured guest will discuss why SEO isn't dead (it just grew up), the biggest mistake agencies are making with AI, how to 10x output without adding headcount, and why your unique data is the unfair advantage that separates you from every other agency prompting ChatGPT and hoping for magic. Terry Zelen is the founder of Zelen Communications, a 35-year-old agency that pivoted aggressively into AI over the last three years. He's helping clients win visibility across both search engines and large language models (LLMs) and even building AI tools internally to reduce hallucinations and improve accuracy. Terry has a degree in marine biology, so marketing wasn't the master plan. After college, he tried breaking into the creative world with zero portfolio and got laughed out of the room; until one person gave him a shot. He worked for free, proved himself, connected with a freelance rep, and slowly worked his way up through the agency ranks. He eventually transitioned from freelancer to agency owner by acquiring his own accounts and building relationships locally in Tampa. Fast forward three decades and now he's helping clients navigate AI, LLM visibility, and what modern SEO really looks like. In this episode, we'll discuss: Why SEO is more complicated now, but agencies willing to adapt can still win How LLM visibility will win you business AI: The greatest leverage small businesses have ever had Building an AI consensus engine Subscribe Apple | Spotify | iHeart Radio Sponsors and Resources This episode is brought to you by Wix Studio: If you're leveling up your team and your client experience, your site builder should keep up too. That's why successful agencies use Wix Studio — built to adapt the way your agency does: AI-powered site mapping, responsive design, flexible workflows, and scalable CMS tools so you spend less on plugins and more on growth. Ready to design faster and smarter? Go to wix.com/studio to get started. SEO Is Not Dead. It's Just Way More Complicated There's a lot of noise right now around "SEO is dead" or "zero-click internet." But that's an oversimplification. SEO isn't going away. It's evolving. Today, it's not just SEO. It's: GEO (Generative Engine Optimization) AEO (Answer Engine Optimization) Local SEO EEAT (Experience, Expertise, Authority, Trust) Search intent In other words, visibility is the game. Not just ranking in Google, but showing up in LLMs like ChatGPT, Gemini, and Perplexity. Terry points out that while snippets and AI-generated summaries are increasing, people still want to verify sources. They're not buying a couch because an LLM told them it's the best. They'll still visit sites, compare options, and validate credibility. Backlinks, structured content, schema, quality. It all still matters. What's different is that now you're playing the game with Google and the LLMs. How LLM Visibility Actually Wins Business This isn't theoretical. Terry shared a story of a client who builds modular classroom buildings. A school district searched for "best mobile building producer in Florida" and the client showed up in a snippet. That visibility led directly to a new contract. So you're no longer optimizing just for rankings. You're optimizing to be the referenced authority when AI generates an answer. That means you better have structured content, clear positioning, backlinks, authority signals, and presence on surfaces LLMs scrape (including platforms like Reddit, though that's evolving). The agencies that understand this shift can bolt on new services like AI SEO or GEO and, in some cases, significantly increase revenue. But there's a catch. This space is evolving fast. What works today might not work next quarter. That's why Terry avoids gray-hat tactics and focuses on fundamentals. AI Is the Greatest Leverage Small Agencies Have Ever Had Terry believes this might be the most exciting time ever for small agencies because AI has eliminated barriers that used to require massive budgets. When a small restaurant client wanted a red snapper on a black background for their website, stock photography didn't cut it and real shoot would've required a diver, photographer, cooperative fish and a significant budget. Instead, they used Midjourney to create the image. Then they animated it so the fins and gills subtly moved. The client was blown away. For a small restaurant, this level of visual production used to be impossible. Now it's affordable and scalable. That's the opportunity. Agencies can deliver higher-quality creative, faster, and at lower cost if they know how to use the tools. A Very Real Fear for Future Marketers Terry regularly speaks to marketing students who are worried AI will take their jobs. What he tells them is that AI won't take your job, but someone who knows how to use AI will. The key is not blind reliance. It's intelligent leverage. AI is excellent at: Research Proposal drafting Competitive analysis First drafts of content Summarizing data What used to take weeks can now take hours. That frees your team from repetitive, dreaded tasks and allows them to focus on strategy, creativity, and client impact. But there's a danger in over-reliance. Too many agencies are slapping "AI" on everything without adding original thinking or proprietary data. Your edge isn't that you use AI. Your edge is your data. Every agency has unique client data, performance metrics, positioning, and experience. When you combine that with AI, that's where real leverage happens. Building a Consensus Engine to Reduce AI Hallucinations One of the more advanced things Terry is experimenting with is what he calls a "consensus engine." The problem with LLMs is that they're probabilistic, not deterministic. Ask the same question twice and you'll get two slightly different answers. They also hallucinate. To combat this, Terry built a workflow using N8N (a Zapier-like automation tool) that runs content through multiple LLMs. One writes it. Another critiques it. The final output must pass both systems before it's considered valid. If they disagree, it's sent back through with adjusted parameters. He's also exploring how different LLMs perform best in different roles: Perplexity for real-time research ChatGPT for writing Claude for programming Instead of treating AI as one tool, he's assembling a stack of specialized tools. That mindset shift, thinking like a systems architect instead of a prompt typist, is what separates surface-level AI use from strategic advantage. Do You Want to Transform Your Agency from a Liability to an Asset? Looking to dig deeper into your agency's potential? Check out our Agency Blueprint. Designed for agency owners like you, our Agency Blueprint helps you uncover growth opportunities, tackle obstacles, and craft a customized blueprint for your agency's success.

Live from Morgan Stanley's TMT conference, our panel break down where AI is already delivering real returns—and where rapid advances are raising new risks.Read more insights from Morgan Stanley.----- Transcript -----Michelle Weaver: Welcome to Thoughts on the Market. I'm Michelle Weaver, U.S. Thematic and Equity Strategist here at Morgan Stanley.Today we've got a special episode on AI adoption. And this is a first in a two-part conversation live from our Technology, Media and Telecom conference.It's Thursday, March 5th at 11am in San Francisco.We're really excited to be here with all of you taping live. And we've got on stage with me. Stephen Byrd, he's our Global Head of Thematic and Sustainability Research; Josh Baer, Software Analyst; and Lindsay Tyler, TMT Credit Research Analyst.So, Stephen, I want to start with you, pretty broad, pretty high level. We recently published our fifth AI Mapping Survey that identifies how different companies are exposed to the broad AI theme. Can you just share with us some insights from that piece and how stocks are performing with this AI exposure?Stephen Byrd: Yeah, it's interesting. I mean, we've been doing this survey now, thanks to you, Michelle, and your excellent work, for quite a while. And every six months it is pretty telling to see the progression.I would say a few things that got my attention from our most recent mapping was the number of companies that are quantifying the adoption benefits continues to go up quite a bit. And to me that feels like that's going to be table stakes very soon as in every industry you see two or three companies that are really laying out quite specifically what they expect to be able to do with AI and lay out the math. I think that really is going to pull all the other companies to follow suit. So, we're seeing that in a big way.We do see adopters, with real tangible benefits performing well. But a new thing that we're seeing now, of course, in the market is concerns that in some cases adoption can lead to dramatic deflation, disruption, et cetera. That's coming up as well. So, we're seeing greater concerns around disruption as well.But broadly, I'd say a proliferation of adoption, that that universe of companies continues to grow, increases in quantification of the benefits. So, that is good. What's really surprised me though, is the narrative among investors has so quickly moved from those benefits which we've talked about into flipping that to toggle all negative, which I know some of our analysts have to deal with every day. The mapping work suggests significant benefits. But the market is fast forwarding to very powerful AI that is very disruptive in deflation. And that's been a surprise to me.Michelle Weaver: Mm-hmm. Josh, I want to bring software into this. Your team has been arguing that AI is actually good for software. And it's really something that you need that application layer to then enable other companies to adopt AI. Can you tell us a little bit about how much GenAI could add to the broader enterprise software market? And how are you thinking about monetization these days?Josh Baer: Of course. I think the best starting place is a reminder that AI is software, and so we see software as a TAM expander. And in many ways, even though this is extremely exciting innovation, it's following past innovation trends where first you see value accrue and market cap accrue to semiconductors, and then hardware and devices, and then eventually software and services. And we do think that that absolutely will occur just given [$]3 trillion in infrastructure investment into data centers and GPUs.There's got to be an application layer that brings all of these productivity and efficiency gains to enterprises and advanced capabilities to consumers as well. And so we see AI more as an evolution for software than a revolution. An evolution of capabilities and expansion of capabilities. LLMs and diffusion engines absolutely unlocked all of these new features of what software can do. But incumbents will play a key role in this unlock.And our CIO surveys really support that. Quarterly we ask chief information officers about their spending intentions, and these application vendors who we cover in the public markets are increasingly selected as vendors that companies will go to, to help deploy and apply AI and LLM technologies.So, to answer your question, we estimate GenAI could unlock [$]400 billion in incremental TAM for software; for enterprise software by 2028. And this is based on looking at the type of work able to be automated, the labor costs associated with that work, the scope of automation, and then thinking about how much of that value is captured typically by software vendors.Michelle Weaver: And you have a bit of a different lens on AI adoption. So, what are some of the ways you're hearing software customers using these AI tools and anything interesting that popped up at the conference?Josh Baer: To echo what Stephen laid out, I mean, all of our software companies are using AI internally, both to drive efficiencies, but also to move faster. So thinking about product. Innovation, you know, the incumbents are able to use all of the same coding tools and, you know, …Michelle Weaver: Mm-hmm.Josh Bear: … products geared to developers to move faster and more efficiently on R&D. So, they're doing more. From a sales and marketing perspective, a G&A perspective, every area of OpEx, our software companies are in a great position to deploy the AI tools internally.I think more important[ly], speaking to this TAM and expanded opportunity, is our companies have skews that they're monetizing. It might be a separate suite that incorporates advanced AI functionality. It might be a standalone offering, or it might be embedded into the core platform because the essence of software is AI and it, you know, leading to better retention rates and acceleration from here.Michelle Weaver: Mm-hmm. And Stephen, going back to you on the state of play for AI, we had the AI labs here and we heard a lot about the developments and what's to come. So, what's your view on the trajectory for LLM advancements and what are some of the key signposts or catalysts you're watching here?Stephen Byrd: Yeah, this is for me, maybe the most important takeaway of the conference – is this continued non-linear improvement of LLMs, which we've been writing about for quite some time. And just to give you an example, we think many of the labs have achieved a step change up in terms of the compute that they have, in some cases 10 x the amount of compute to train their LLMs. And that [if] the scaling laws hold – and we see every sign that they will – a 10x increase in compute used to train the models results in about a doubling of the model capabilities.Now just let that sink in for a moment. Let's just think about that. A doubling from here in a relatively short period of time is difficult to predict. It's obviously very significant and I think several of the LLM execs at our event sounded to me extremely bullish on what that will be. A lot of that I think will be evident in greater agentic capabilities.But also, I'd say greater creativity. It was about three weeks ago, three of the best physics minds in the world worked with an LLM to achieve a true breakthrough in physics – solving a problem that had never been solved before. A couple of days ago, a math team did the same thing. And so, what we're seeing is sort of these breakthrough capabilities in creativity. This morning I thought Sam speaking to, you know, incredible increases in what these models can do – which also brings risk. You know, I think it was interesting he spoke to, you know, the risk of misalignment, the risk of what these models are doing.But for me, that's the single biggest thing that I'm thinking about, and that's going to be evident in the next several months.Michelle Weaver: Mm-hmm.Stephen Byrd: So, you know, on the positive side, it leads to greater benefits from AI adoption. And to Josh's point that, you know – more and more the economy can be addressed by AI, I do get concerned about the risk that that kind of step change will create greater concerns about disruption and deflation.That causes me to think a lot about that dynamic. Interestingly, we think the Chinese labs will not be able to keep pace just for one reason, which is compute. We think the Chinese labs have everything else they need. They have the talent, the infrastructure. They certainly have the energy, power. But they don't have the chips.If what we laid out with the American models turns out to be true, I could see a chain reaction where the Chinese government pushes the Trump administration for full transfer of the best technology to China. And China could use their rare earth trade position to ensure that. So, that's sort of the chain reaction I've been thinking about.Michelle Weaver: Mm-hmm. So, let's think about then bottlenecks in the U.S. Power is still one of the main bottlenecks. We had several of the solutions providers here at the conference. So, what are you thinking in terms of the size of the power bottleneck in the U.S. and how are we going to fix that?Stephen Byrd: Yeah, absolutely. I am bullish on the companies that can de-bottleneck power, not just in the U.S., a few other places. Let's go through the math in terms of the problem we face and then the solution.So, we have this very cool – it is cool if you're a nerd – power model that starts in the chip level up, from our semiconductor teams. And from that, we build a global power demand model for data centers. We then apply that to the U.S.Through 2028 we need about 74 gigawatts of data centers, both AI and non-AI to be built in the United States. I don't think we'll be able to achieve that for lots of reasons. But starting from that 74, we have sort of 10 gigs that have been recently built or are under construction. We have 15 gigs of incremental grid access, but after those two, we have to go to unconventional solutions, meaning typically off-grid solutions, over 40 gigawatts of unconventional solutions.So that will be repurposing Bitcoin sites, which could be sort of 10 to 15 gigawatts. That'll be big. Renewable energy, fuel cells will be part of the solution. Gas turbines will be a big part of the solution. Co-locating at a few nuclear plants. I'm less bullish than I used to be on that. But when we net all that out, we think the U.S. is likely to be 10 to 20 percent short of the data center capacity that will need to be in.It's not just a power grid access issue, though, that's a big one. Labor is now showing up as a huge issue. Many of the companies I speak to trying to develop data centers struggle with availability of labor. Electricians being one very tangible example. In the U.S. we need hundreds of thousands of additional electricians.So, for any of your children, like mine, thinking about careers, you know, you'd be surprised [at] the amount of money that people are making in the infrastructure business that does feel like it's a labor shift that's going to have to happen, but it's going to take years. So, in that context, we had a number of the Bitcoin companies at our event here. And the economics of turning a Bitcoin site into hosting a data center are extremely attractive. I mean, extremely attractive.To give you a sense of that. Before this opportunity presented itself to these Bitcoin players, those stocks tended to trade at an enterprise value per watt of about $1 to $2 a watt. Then we started to see these deals in which the Bitcoin players build a data center and lease them to hyperscalers. Those deals – depends a lot on the deal but – have created between $10 and $18 a watt of value. Let me repeat that. 10 to 18 – relative to where these stocks were at 1 to 2.Now many of these stocks have rerated, but not all of them. And there's still quite a bit of upside. And what we've noticed is the economics that the hyperscalers are paying are trending up and up and up. Because of this power shortage that we're dealing with. So, a lot of exciting opportunities are still in the power space.Michelle Weaver: Great. Well, I think that's a good place to wrap this first part of our conversation around AI adoption and the state of play. We'll be back again tomorrow with Part Two, looking at financing and risks.To our panelists, thank you for talking with me. And to our audience, thanks for listening. If you enjoy Thoughts on the Market, please leave us a review wherever you listen and share the podcast with a friend or colleague today.

AI Reality Check: Did the LLM Job Apocalypse Begin Last Week? Cal Newport takes a closer look at recent AI news. Below are the topics covered in today's episode (with their timestamps).  Get your questions answered by Cal! Here's the link: https://bit.ly/3U3sTvo Video from today's episode:youtube.com/calnewportmedia STORY #1: Jack Dorsey announces layoffs at Block [1:28] STORY #2: The education level of LLM-based tools [11:45] STORY #3: What's happening in the world of computer programming? [19:24] Links: Buy Cal's latest book, “Slow Productivity” at www.calnewport.com/slow  Get a signed copy of Cal's “Slow Productivity” at https://peoplesbooktakoma.com/event/cal-newport/ https://x.com/jack/status/2027129697092731343 https://www.nytimes.com/2026/02/26/technology/block-square-job-cuts-ai.html https://x.com/emollick/status/2027153371241607420 https://www.forbes.com/sites/ronshevlin/2026/02/27/block-lays-off-40-of-staff-and-blames-it-on-ai-dont-buy-the-excuse/ https://www.youtube.com/watch?v=56HJQm5nb0U http://calnewport.com Thanks to Jesse Miller for production and mastering.   Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Meta Ray-Bans are sending private videos to human workers in Kenya, and Dr. Niki talks about what we know about the effects of LLM use on mental health.Starring Jason Howell, Tom Merritt, and Dr. Niki.Links to stories discussed in this episode can be found here. Hosted on Acast. See acast.com/privacy for more information.

This week we talk about Anthropic, the Department of Defense, and OpenAI.We also discuss red lines, contracts, and lethal autonomous systems.Recommended Book: Empire of AI by Karen HaoTranscriptLethal autonomous weapons, often called lethal autonomous systems, autonomous weapons systems, or just ‘killer robots,' are military hardware that can operate independent of human control, searching for and engaging with targets based on their programming and thus not needing a human being to point it at things or pull the trigger.The specific nature and capabilities of these devices vary substantially from context to content, and even between scholars writing on the subject, but in general these are systems—be they aerial drones, heavy gun emplacements, some kind of mobile rocket launcher, or a human- or dog-shaped robot—that are capable of carrying out tasks and achieving goals without needing constant attention from a human operator.That's a stark contrast with drones that require either a human controlled or what's called a human-in-the-loop in order to make decisions. Some drones and other robots and weapons require full hands-on control, with a human steering them, pointing their weapons, and pulling the trigger, while others are semi-autonomous in that they can be told to patrol a given area and look for specific things, but then they reach out to a human-in-the-loop to make final decisions about whatever they want to do, including and especially weapon-related things; a human has to be the one to drop the bomb or fire the gun in most cases, today.Fully autonomous weapon systems, without a human in the loop, are far less common at this point, in part because it's difficult to create a system so capable that it doesn't require human intervention at times, but also because it's truly dangerous to create such a device.Modern artificial intelligence systems are incredibly powerful, but they still make mistakes, and just as an LLM-based chatbot might muddle its words or add extra fingers to a made-up person in an image it generates, or a step further, might fabricate research referenced in a paper it produces, an AI-controlled weapon system might see targets where there are no targets, or might flag a friendly, someone on its side, or a peaceful, noncombatant human, as a target. And if there's no human-in-the-loop to check the AI's understanding and correct it, that could mean a lot of non-targets being treated like targets, their lives ended by killer robots that gun them down or launch a missile at their home.On a larger scale, AI systems controlling arrays of weapons, or even entire militaries, becoming strategic commanders, could wipe out all human life by sparking a nuclear war.A recent study conducted at King's College London found that in simulated crises, across 21 scenarios, AI systems which thought they had control of nation-state-scale militaries opted for nuclear signaling, escalation, and tactical nuclear weapon use 95% of the time, never once across all simulations choosing to use one of the eight de-escalatory options that were made available to them.All of which suggests to the researchers behind this study that the norm, approaching the level of taboo, associated with nuclear weapons use globally since WWII, among humans at least, may not have carried over to these AI systems, and full-blown nuclear conflict may thus become more likely under AI-driven military conditions.What I'd like to talk about today is a recent confrontation between one AI company—Anthropic—and its client, the US Department of Defense, and the seeming implications of both this conflict, and what happened as a result.—In late-2024, the US Department of Defense—which by the way is still the official title, despite the President calling it the Department of War, since only Congress can change its name—the US DoD partnered with Anthropic to get a version of its Claude LLM-based AI model that could be used by the Pentagon.Anthropic worked with Palantir, which is a data-aggregation and surveillance company, basically, run by Peter Thiel and very favored by this administration, and Amazon Web Services, to make that Claude-for-the-US-military relationship happen, those interconnections allowing this version of the model to be used for classified missions.Anthropic received a $200 million contract with the Department of Defense in mid-2025, as did a slew of other US-based AI companies, including Google, xAI, and OpenAI. But while the Pentagon has been funding a bunch of US-based AI companies for this utility, only Claude was reportedly used during the early 2026 raid on Venezuela, during which now-former Venezuelan President Maduro was taken by US forces.Word on the street is that Claude is the only model that the Pentagon has found truly useful for these sorts of operations, though publicly they're saying that investments in all of these models have borne fruit, at least to some degree.So Anthropic's Claude model is being used for classified, military and intelligence purposes by the US government. Anthropic has been happy about this, by all accounts, because that's a fair bit of money, but also being used for these purposes by a government is a pretty big deal—if it's good enough for the US military, after all, many CEOs will see that as a strong indication that Claude is definitely good enough for their intended business purposes.On February 24 of 2026, though, the US Defense Secretary, Pete Hegseth, threatened to remove Anthropic from the DoD's stable of AI systems that they use unless the company allowed the DoD to use Claude for any and all legal purposes—unrestricted use of the model, basically.This threat came with a timeline—accede to these demands by February 27 or be cut from the DoD's supply chain—and the day before that deadline, the 26th, Anthropic's CEO released a statement indicating that the company would not get rid of its red lines that delineated what Claude could and could not be used for, and on the 27th, US President Trump ordered that all US agencies stop using Anthropic tools, and said that he would declare the company a supply chain risk, which would make it illegal for any company doing business with the US government at any level and in any fashion to use Anthropic products or services—a label that's rarely used, and which was previously used by the Trump administration against Chinese tech giant Huawei on the basis that the company might insert spy equipment in communications hardware installed across the US if they were allowed to continue operating in the country.Those red lines that Anthropic's CEO said he wouldn't get rid of, not even for a client as big and important as the US government, and not even in the face of threats by Hegseth, including that he might invoke the Defense Production Act, which would allow him to force the company to allow the Pentagon to use Claude however they like, or Trumps threat that the company be blacklisted from not just the government, but from working with a significant chunk of Fortune 500 companies, those red lines include not allowing Claude to be used for controlling autonomous weapon systems, killer robots, basically, and not allowing Claude to be used for surveilling US citizens.The Pentagon signed a contract with Anthropic in which they agreed to these terms, but Hegseth's new demand was that Anthropic sign a new version of the contract in which they allow the US government to use Claude and their other offerings for ‘all legal purposes,' which apparently includes, at least in some cases and contexts, killer robots and mass surveillance.So the Pentagon tried to strong-arm a US-based AI company into allowing them to use their product for purposes the company doesn't consider to be moral, and that led to this situation in which Anthropic is now being phased out from US government use—it'll apparently take about 6 months to do this, and some analysts speculate that timeline is meant to serve as a period in which further negotiation can occur—but either way, it's being phased out and it may even have trouble getting major clients in the future as a result of being blackballed.As all this was happening, OpenAI stepped in and offered its products and services to fill the void left by Anthropic in the US government.OpenAI's CEO has been cozying up to Trump a lot since he regained office, and has positioned the company as a major US asset, too big to fail because then China will win the AI race, basically, so this makes sense. Its CEO released several statements and press releases in the wake of this further cozying, saying that they believe the same things Anthropic does, and that they're not giving up any credibility for doing this because they have the same red lines, no killer robots, no mass surveillance of US citizens.But this is generally assumed to be bunk, because why would the Pentagon agree to the same terms all over again, and with a company that provides, for their purposes and right now, anyway, inferior services instead of the one they just chased out and blackballed, and which was helping them do purposeful, effective things, like kidnapping a foreign leader from a secure facility, today?Instead, what it sounds like is OpenAI is trying to have its cake and eat it too, saying publicly that they don't want their offerings used to control autonomous weapons systems or mass surveil Americans, but instead of writing that into the contract, they've got some basic guardrails baked into their systems, and they are assuming those guardrails will keep any funny business from happening. So it's a sort of gentleman's agreement with their clients that OpenAI products won't be used for mass surveillance or killer robots, rather than something legally binding, as was the case with Anthropic.The response to all this within the tech world has been illustrative of what we might expect in the coming years. Many people, including folks working on these technologies, are halting their use of OpenAI tech in protest, and in some (at this point at least) fewer cases, people are quitting their OpenAI jobs, because they are strongly opposed to these use-cases and would prefer to support a company that takes a strong stand on these sorts of moral issues.Some analysts also wonder if this will ensure the Pentagon only ever has access to inferior AI models because they intentionally threatened and disempowered a key AI industry CEO in public, saying that they had final say over how these tools are used, and many such CEOs are both unaccustomed to such stripping down, but are also doing the work they're doing for ideological reasons—they have beliefs about what the future, as enabled by AI technologies, will look like, and they believe they will play a vital role in making that future happen.The idea, then, is why would they want to work with the Pentagon, or the US government more broadly, if that means no longer being in charge of the destiny of these tools they're putting so much time, effort, and resources into building? Why would they take on a client, even a big, important one, if that means no longer having any grain of control over the future of the world as shaped by the systems they're building?We'll know a bit more about how all this plays out within the next handful of months, as this could serve as a moral differentiator between otherwise near-match products in the AI category, allowing companies like Anthropic to compete, both in terms of clients and in terms of employees, with the likes of OpenAI and xAI by saying, look, we don't want killer robots or mass surveillance and we gave up a LOT, put our money where our mouths are, in support of that moral stance.That could prove to be a serious feather in their cap, despite the initial cost, though it could also be that the pressure the US government is willing and able to apply to them instead serves as a warning to others, and the likes of OpenAI and Google and so on just get better at speaking out of both sides of their mouths on this issue, creating sneakier contracts that allow them to say the same on paper, seeming to take the same moral stance Anthropic did, while behind closed doors allowing their clients to do basically whatever they want with their products, including using them to control killer robots and to mass surveil US citizens.Show Noteshttps://www.kcl.ac.uk/news/artificial-intelligence-under-nuclear-pressure-first-large-scale-kings-study-reveals-how-ai-models-reason-and-escalate-under-crisishttps://www.axios.com/2026/02/26/ai-nuclear-weapons-war-pentagon-scenarioshttps://www.nytimes.com/2026/02/27/technology/openai-agreement-pentagon-ai.htmlhttps://en.wikipedia.org/wiki/Lethal_autonomous_weaponhttps://www.theverge.com/ai-artificial-intelligence/885963/anthropic-dod-pentagon-tech-workers-ai-labs-reacthttps://www.theverge.com/ai-artificial-intelligence/886816/openai-reached-a-new-agreement-with-the-pentagonhttps://arstechnica.com/tech-policy/2026/02/trump-moves-to-ban-anthropic-from-the-us-government/https://apnews.com/article/anthropic-pentagon-ai-dario-amodei-hegseth-0c464a054359b9fdc80cf18b0d4f690chttps://www.wsj.com/tech/ai/whats-really-at-stake-in-the-fight-between-anthropic-and-the-pentagon-d450c1a1https://openai.com/index/our-agreement-with-the-department-of-war/https://www.kcl.ac.uk/news/artificial-intelligence-under-nuclear-pressure-first-large-scale-kings-study-reveals-how-ai-models-reason-and-escalate-under-crisishttps://www.axios.com/2026/02/26/ai-nuclear-weapons-war-pentagon-scenarios This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

Starting off in FOLLOW UP, we've got a tax economist who actually made money betting against the "efficiency" of Elon's budget-slashing fever dreams, while Tesla is busy trying to dodge a $243 million jury verdict for an Autopilot-assisted fatality. Not content with being legally liable, Tesla is also suing the California DMV because they're offended someone called their "Autopilot" and "Full Self-Driving" marketing deceptive—ironic, since Jack Dorsey just "proactively" halved the staff at Block to make room for more AI slop. Speaking of which, Goldman Sachs is here to remind us that all this AI spending added a grand total of zero to the US GDP last year, mostly because we're just exporting all that cash to overseas chip makers while 80% of execs admit the tech hasn't actually done anything for productivity yet.Moving into IN THE NEWS, Sam Altman had the audacity to compare ChatGPT's energy-sucking habits to the 20-year evolution of a human, though the internet wasn't exactly buying the "my bot is just like a baby" defense. Anthropic actually stood its ground against the Pentagon's demand for killer robots and mass surveillance, so naturally, the military just signed a deal to put Elon's Grok in their classified systems instead—because what could go wrong with an "edgy" LLM in the war room? Meanwhile, cities are dumping AI surveillance contracts as citizens start a literal "smash-the-snitch-box" campaign against Flock's license plate readers, Google's AI is busy inserting racial slurs into news alerts, and the White House is apparently harboring a staffer moonlighting as a racist "masterpiece" creator on X. We've also got Reddit being slapped with a $20 million fine in the UK for being lazy with age checks, while Discord and Apple scramble to build verification tools that hopefully won't leak your entire identity to a hacker in Belarus.In MEDIA CANDY, the Paramount-Skydance merger is leaving the industry in a cold sweat of "synergy" layoffs, but at least we're getting more Game of Thrones spinoffs and Star Trek reboots to rot our brains. Face/Off 2 lost its director, Ryan Coogler is taking on The X-Files, and Google wants to use AI to turn music into generic "lo-fi" background noise for the masses.Over in APPS & DOODADS, OpenAI is planning a 2027 smart speaker that literally watches you through a camera—because you definitely wanted a $300 Sam Altman-shaped eye in your kitchen—while the Dark Sky creators are back with "Acme Weather" for the low price of $25 a year.We wrap up THE DARK SIDE WITH DAVE with a deep dive into "Under Pressure" and Coruscant's urban sprawl, leaving us to reminisce about the days when KPT Bryce was the pinnacle of tech—back when "generative art" was just a fractal that took six hours to render.Sponsors:DeleteMe - Get 20% off your DeleteMe plan when you go to JoinDeleteMe.com/GOG and use promo code GOG at checkout.SquareSpace - go to squarespace.com/GRUMPY for a free trial. And when you're ready to launch, use code GRUMPY to save 10% off your first purchase of a website or domain.Private Internet Access - Go to GOG.Show/vpn and sign up today. For a limited time only, you can get OUR favorite VPN for as little as $2.03 a month.SetApp - With a single monthly subscription you get 240+ apps for your Mac. Go to SetApp and get started today!!!1Password - Get a great deal on the only password manager recommended by Grumpy Old Geeks! gog.show/1passwordShow notes at https://gog.show/735Watch on YouTube: https://youtu.be/jdz--v3eeU4FOLLOW UPGuy Bets Entire Life Savings Against Elon Musk, WinsTesla sues California DMV after it banned the term 'Autopilot'Jack Dorsey just halved the size of Block's employee base — and he says your company is nextIN THE NEWSSam Altman: Know What Else Used a Lot of Energy? Human CivilizationStatement from Dario Amodei on our discussions with the Department of WarAnthropic Tells Pete Hegseth to Take a HikeCities Are Shredding Their AI Surveillance Contracts en MasseKalshi Suspended a California Politician and a YouTuber for Insider TradingDiscord delays age verification to address user concernsApple introduces age verification for apps in Utah, Louisiana and AustraliaMEDIA CANDYAs Paramount Skydance wins the battle for Warner Bros. as Netflix ends its bid, here's the mood inside all three companies.A Knight of the Seven KingdomsStar Trek: Starfleet AcademyThe Night Agent Season 3'Face/Off 2' Director Adam Wingard is Now/GoneRyan Coogler's X-Files reboot gets the green light at HuluMortal Kombat II | Official Trailer IIGoogle's AI Slop Machine Is Coming for Your MusicDropping Names... and other things with Jonathan Frakes and Brent SpinerOnce We Were SpacemenAPPS & DOODADSOpenAI will reportedly release an AI-powered smart speaker in 2027Instagram Will Notify Parents When Teens Use Search Terms Related to SuicideThe creators of Dark Sky have a new weather appThis App Warns You if Someone Is Wearing Smart Glasses NearbyTHE DARK SIDE WITH DAVEDave BittnerThe CyberWireHacking HumansCaveatControl LoopOnly Malware in the BuildingStrong Songs - S08E02 - "Under Pressure" by Queen and David BowieThe Problem with Coruscant (Planet Cities Explained)Reminds me of KPT Fractal ExplorerKPT Bryce 1.0 with John Dvorak and Kai KrauseSingle-Biome PlanetKPT Shapes by Dave BittnerBald Mr Clean mascot "retired"My childhood disappointment with scrubbing bubbles.CLOSING SHOUT-OUTSActor Robert Carradine Dies At Age 71See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.