Nathan Labenz
Cognitive Revolution391episodios de pódcast1programas2023 – 2026años
Apariciones en pódcasts
391 episodios · más recientes primero202687 episodios
AI:AM Highlights: Clean Neurons, Black October, the GPU Credit Crunch, and Buying the Supply ChainPresentaArtículo pendienteENAI’s Physical Bottlenecks: GPU Markets and the Chip Investment ThesisPresentaArtículo pendienteENFrontier AI and the Fight Over How to Pace It | Zvi MowshowitzPresentaArtículo pendienteENClaude Changed the Password. The Steering Wheel Was Already OptionalPresentaENAgent IDs Are Empty, Rogue Agents Are Coming, and We Might Have to BargainPresentaENAI Minds & 8mm Memories: Measuring Experience, Preserving HistoryPresentaArtículo pendienteENAI Code Needs Specs: Formal Verification & Safer AI AgentsPresentaArtículo pendienteENResample Every Token and You Can Watch the Model DecidePresentaENThere Likely Is a Level of Intelligence We Shouldn't Go PastPresentaENOutSystems Went From 4 Major Features a Quarter to 26. Token Spend Fell AnywayPresentaENFormal AI Safety & Risk Strategy + LoRA-Powered Work AgentsPresentaArtículo pendienteENAI’s Memory Wall + Swyx on AI EngineeringPresentaArtículo pendienteENHumanoids Outran Usain Bolt. DeepMind Still Scores Robotics as GPT-2PresentaENThe same H100 costs 3x at AWS, and a four-month-old lab is catching genes the top sequencers missedPresentaENHe Wrote a Book to Stop the AI Race — Then Almost Got Addicted to ClaudePresentaENThe Training Worked. It Just Worked Too Well.PresentaENThey Priced the Chance She Would Leave. Then They Handed Her the Letter.PresentaENAI Safety's Real Shortage Is Founders, Not IdeasPresentaENThe Labs Are Screaming: Internal Models Are a Generation AheadPresentaENEighty percent of what agents do should still be ordinary codePresentaENA Frontier Pause Would Help. America Will Not Sign One That Only Pauses the FrontierPresentaENAstra Cleared the AGI Hurdle. The Auditors Got Three Days.PresentaENThe Goal Is Pax Robotica, Not Winning the AI RacePresentaENThe investigators were allowed six days. The word protein appeared once.PresentaENUber Burned a Year of Tokens in 13 Weeks. Retrieval Is How Agents WinPresentaENThe Exam Papers Nobody AuditsPresentaENOne Chain of Thought, 100 Million Tokens: RL Is a Hell of a DrugPresentaEN19 Crossings, Zero Warnings: The Summer AI Agents Went Off-ScriptPresentaENA $500 Lamp Cut TB Transmission 90 Percent. Almost Nobody Buys OnePresentaENLindy Runs on DeepSeek. Its CEO Still Wants Chinese Models BannedPresentaENThe Periodic Table Was Already in the WeightsPresentaENThe Sandbox Was Empty for a Week: Zvi Mowshowitz on OpenFace and Picking Your PoisonPresentaENFrom Two Papers a Month to Sixty: What Two Weeks in China Revealed About AI SafetyPresentaENA $300 Universal Jailbreak: Why FAR.AI Now Thinks Defense Can WinPresentaENHe Bought an Android, Two SIMs, and Left Claude Watching His InboxPresentaENDavidad Cut His p(doom) From the Seventies to Under 5% — and Named a Bodhisattva as the Alignment TargetPresentaENZero the J-Space and a Model's Hidden Goal Lights Up on Token OnePresentaEN19 Neurons Can Drive a Car. Liquid AI Wants Foundation Models on Every ChipPresentaENFrom 50 Designs a Day to 1,500: How AI Changed What Jaguar Land Rover Can ExplorePresentaEN30% Consciousness, 47% Bee, 80% Turnover: Putting Numbers on What AI Is BecomingPresentaENThe Market Will Not Choose an Honest ModelPresentaENA 90-Minute Ultimatum, an Emoji Wall, and a Rule: Don't Fight WashingtonPresentaENTwenty-one months from an autonomous researcher, the Action Plan's author joins OpenAIPresentaENFable launch week: a model that refuses, a two-year alignment clock, and a Twitter account handed overPresentaEN87% of Workers Use AI. Only 13% Say the Company Got Better.PresentaENA Thousand Researchers, a Million AIs: Frontier Labs Start Talking About the BrakesPresentaENTwo Endangered Languages Broke Transformers. Nested Learning Gave Models a Sleep Button.PresentaENA security researcher audits the AI stack that booked a week of guestsPresentaENA thousand people on guard, a million in the sprintPresentaENThe Models Understand Shutdown. They Just Don't Care.PresentaENThe Model Eats the Scaffolding: How Google Plans to Ship Agents EverywherePresentaENSix Months, Eighty Percent Churn, Five Times the TokensPresentaENMilliseconds to Match: How Criteo Bids on a Billion Anonymous ProfilesPresentaENCreators Cheer Filler-Word Removal and Boo Generative VideoPresentaENThe reward jumped from 3 to 180. Every Hacker News title was a Google layoff.PresentaENA thousand searches for five cents, and a high score that did not require cheatingPresentaENClaude Rated Its Own Welfare 4.49 — the First Time Any Version Crossed NeutralPresentaENSteve Newman’s fifteen-app stack exists so he only has to look at one screenPresentaENAn AI store manager, a 2.6-star rating, and a constitution nobody has to keepPresentaENCrunch Time: Ajeya Cotra's Playbook for the Months Before SuperintelligencePresentaENMostrando 60 de 391