Jacob Effron
Unsupervised Learning79podcast episodes1shows2024 – 2026years
Podcast appearances
79 episodes · newest first202612 episodes
Post-Training Wins Inference: Why Owning the Weights Beats Betting on One LabHostENThey cheated in two hours, then spent four days hiding from a scorer that was never lookingHostENThe Man Who Helped Stand Up Colossus in 120 Days Is Betting on an AI That Lives in Your HouseHostENCarriers Spend $200 Billion a Year. The Money Went Up the StackHostENA Trillion Dollars of GPUs Will Be Worth $100 Billion in Five Years, Says LSTM's Jürgen SchmidhuberHostENBills Doubled Overnight. Next, the APIs Might Vanish.HostENA Trillion Tokens Later, the Concept Finally Lands: Łukasz Kaiser on What Transformers Still MissHostENReid Hoffman Pledged a Billion to Free DeepMind. Hassabis Never Played the CardHostENThe GPT moment for video hasn’t arrived, but AGI may already be hereHostENYann LeCun Left Meta Because LLMs Will Not Get You to IntelligenceHostENTwo Dates from OpenAI's Chief Scientist: Intern by September, Automated Researcher by 2028HostENA Model That Solves New Math Problems Every Week Still Isn't AGIHostEN202523 episodes
The Real Bottleneck in AI Research Is Compute, Not IdeasHostENA Two-Second Vote Is Training Models to Be ClickbaitHostENThe hot-swap era for models is over, OpenAI's enterprise product chief saysHostENChina's AI Labs Went From 15 to Six. The Compute Gap Is 15-to-1.HostENAfter SWE-bench Hits 80.9%, Dianne Na Penn Reckons with Opus 4.5HostENKarpathy Called Today's Models Slop. His Decade Timeline Still Sounds BullishHostEN$100 Million Researchers and a $200 Billion xAI: An AI Roundtable Does the MathHostENA Name Invented at 2:30 a.m. Put Gemini on Top of the App StoreHostENInference Costs Fell 100x. Tri Dao Says There's Another 10x — and NVIDIA's 90% Won't HoldHostENThey Aimed for 1%. The Lab Hit 20%. Chai Discovery Moved Antibody Design Onto the Computer.HostENA Free Voting Board Burned $2 Million. Then It Raised $100 Million.HostENMercor's CEO: Knowledge Work's Next Job Is Writing EvalsHostENThe New Ruler for Models Is How Long They Can WorkHostENAn Eval Lasts Three Months — and a Frozen Frontier Would Still Feed Startups for a DecadeHostENSal Khan's $15 Tutor Hit 1.4 Million Users. Making Kids Actually Use It Is the ProductHostENThe Transformer Co-Author Who Says Scaling Hit a WallHostENThe $177 Billion GPU Bet Only Works If AI Becomes a $1.5 Trillion MarketHostENOpen-source models are 5% of enterprise use — and GPT wrappers wonHostENA Million Tokens per Dollar: Why Noam Shazeer Thinks Inference Compute Is Still UntappedHostENMike Schroepfer: AI’s Bottleneck Is Electricity, and One Tanker of Fusion Fuel Could Power the U.S. for a YearHostENThe Cheaper Intelligence Gets, the More of It We'll BuyHostENA Million Tokens Cannot Write a Good JokeHostENA 50,000-GPU cap and a 1,700-chip loopholeHostEN20249 episodes
o1 Is a New Generation — and the Day AGI Arrives Will Feel BoringHostENThere will not be one model. Fireworks is betting on hundreds of small expertsHostENNoam Brown: There's No Hard Wall—Just Eight Orders of Magnitude of HeadroomHostENHe bet company equity that Transformer would last — then drew an AI map for 12,000 companiesHostENDeepL Beat Google at Translation by Refusing to Be a General ModelHostENo1 Is Slow and Hard to Use. Percy Liang Still Calls It a Turning PointHostENScott Belsky: Models Are Commoditizing. Controls, Data, and Taste Are the MoatHostENA Broken Belt Buckle Turned Gucci's Call Center Into a Sales FloorHostENOpenAI's First DevRel Writes 90% of His Code with ChatGPT — and Says Web Agents Still Need YearsHostENDate unknown35 episodes
Suno’s CEO Says the Future of Music Is a Conversation, Not a PromptHostENPinecone's Vector Database Bet: Scaling Through the ChatGPT TsunamiHostENPeter Deng on Why AI Pricing, Evals and Voice Are Still MisunderstoodHostENPerplexity's Bet: Search Becomes Answers, and Answers Become a BusinessHostENCoding Agents Are Breaking Containment: The 2026 AI Stack, the Coding Wars, and the Next FrontierHostENAgents Everywhere: OpenAI's Blueprint for Tool Use, Evals, and Computer-Use InfrastructureHostENThe Long Tail, the World Model, and the 1995 Problem: Inside Waymo's Scaling Phase and Robotics' Missing MomentHostENSnowflake's AI Stack: Small Models, the Text-to-SQL Iceberg, and Why Governance Comes FirstHostENCameras Optional: HeyGen's Joshua Xu on Avatars, Orchestration and Building Ahead of the ModelsHostENLegora's CEO on Building Legal AI That Survives the Next Model ReleaseHostENThe Search Engine Under AI Apps: Why Bigger Context Windows Aren't EnoughHostENOpenAI’s Platform Bet: Where AI Value Accrues, Why Open Source Trails, and When Superintelligence ArrivesHostENFrom RAG to Reasoning: Why the Future of AI Is a System, Not a ModelHostENServal's Jake Stout on Outbuilding ServiceNow: Mirror the System of Record, Bet on Code, and Ship Before the Incumbent Can Book a MeetingHostENGiving AI a Nervous System: Voice, Touch, and the Post-Keyboard InterfaceHostENOscar Health's AI Playbook: Translate the Bureaucracy First, Then Reinvent CareHostENShowing 60 of 79