JE

Jacob Effron

Unsupervised Learning
79épisodes de podcast1émissions2024 – 2026années

Passages en podcast

79 épisodes · plus récents d’abord
Titres en
Pas encore de titres en français : les titres anglais d’origine sont affichés.
202612 épisodes
Unsupervised Learningoct. 2026Post-Training Wins Inference: Why Owning the Weights Beats Betting on One LabModèles et entraînementPost-trainingReinforcement LearningAnimeENUnsupervised Learningsept. 2026They cheated in two hours, then spent four days hiding from a scorer that was never lookingSécuritéAI securitySécurité et gouvernanceAnimeENUnsupervised Learningjuil. 2026The Man Who Helped Stand Up Colossus in 120 Days Is Betting on an AI That Lives in Your HouseAgentsAI startups & small teamsCompany BuildingAnimeENUnsupervised Learningjuil. 2026Carriers Spend $200 Billion a Year. The Money Went Up the StackEntreprise et stratégieAI StrategyIA en entrepriseAnimeENUnsupervised Learningjuil. 2026A Trillion Dollars of GPUs Will Be Worth $100 Billion in Five Years, Says LSTM's Jürgen SchmidhuberModèles et entraînementAGIAI researchAnimeENUnsupervised Learningjuin 2026Bills Doubled Overnight. Next, the APIs Might Vanish.Entreprise et stratégieAI StrategyOpen-source ecosystemsAnimeENUnsupervised Learningjuin 2026A Trillion Tokens Later, the Concept Finally Lands: Łukasz Kaiser on What Transformers Still MissModèles et entraînementAI researchAgents de codeAnimeENUnsupervised Learningjuin 2026Reid Hoffman Pledged a Billion to Free DeepMind. Hassabis Never Played the CardEntreprise et stratégieAI StrategySécurité et gouvernanceAnimeENUnsupervised Learningmai 2026The GPT moment for video hasn’t arrived, but AGI may already be hereModèles et entraînementWorld modelsReinforcement LearningAnimeENUnsupervised Learningmai 2026Yann LeCun Left Meta Because LLMs Will Not Get You to IntelligenceModèles et entraînementWorld modelsAI researchAnimeENUnsupervised Learningavr. 2026Two Dates from OpenAI's Chief Scientist: Intern by September, Automated Researcher by 2028Modèles et entraînementContinual learningReinforcement LearningAnimeENUnsupervised Learningjanv. 2026A Model That Solves New Math Problems Every Week Still Isn't AGIModèles et entraînementReinforcement LearningAGI TimelinesAnimeEN
202523 épisodes
Unsupervised Learningdéc. 2025The Real Bottleneck in AI Research Is Compute, Not IdeasEntreprise et stratégieAI StrategyScaling LawsAnimeENUnsupervised Learningdéc. 2025A Two-Second Vote Is Training Models to Be ClickbaitÉvaluations et observabilitéRL environments & simulatorsData quality & curationAnimeENUnsupervised Learningdéc. 2025The hot-swap era for models is over, OpenAI's enterprise product chief saysEntreprise et stratégieIA en entrepriseProduct StrategyAnimeENUnsupervised Learningdéc. 2025China's AI Labs Went From 15 to Six. The Compute Gap Is 15-to-1.Entreprise et stratégieGeopoliticsAI StrategyAnimeENUnsupervised Learningdéc. 2025After SWE-bench Hits 80.9%, Dianne Na Penn Reckons with Opus 4.5AgentsAI Product ManagementÉvaluationsAnimeENUnsupervised Learningoct. 2025Karpathy Called Today's Models Slop. His Decade Timeline Still Sounds BullishEntreprise et stratégieAI StrategyCompany BuildingAnimeENUnsupervised Learningsept. 2025$100 Million Researchers and a $200 Billion xAI: An AI Roundtable Does the MathEntreprise et stratégieAI StrategyAI startups & small teamsAnimeENUnsupervised Learningsept. 2025A Name Invented at 2:30 a.m. Put Gemini on Top of the App StoreMultimodal et médiasMultimodal AIGenerative mediaAnimeENUnsupervised Learningsept. 2025Inference Costs Fell 100x. Tri Dao Says There's Another 10x — and NVIDIA's 90% Won't HoldInfrastructure et déploiementInferenceAI cost & performanceAnimeENUnsupervised Learningaoût 2025They Aimed for 1%. The Lab Hit 20%. Chai Discovery Moved Antibody Design Onto the Computer.Modèles et entraînementAI for scienceComputational BiologyAnimeENUnsupervised Learningjuin 2025A Free Voting Board Burned $2 Million. Then It Raised $100 Million.Évaluations et observabilitéÉvaluationsBenchmarksAnimeENUnsupervised Learningjuin 2025Mercor's CEO: Knowledge Work's Next Job Is Writing EvalsÉvaluations et observabilitéÉvaluationsData quality & curationAnimeENUnsupervised Learningmai 2025The New Ruler for Models Is How Long They Can WorkAgents de codeIngénierie d’agentsÉvaluationsAnimeENUnsupervised Learningmai 2025An Eval Lasts Three Months — and a Frozen Frontier Would Still Feed Startups for a DecadeModèles et entraînementÉvaluationsPost-trainingAnimeENUnsupervised Learningavr. 2025Sal Khan's $15 Tutor Hit 1.4 Million Users. Making Kids Actually Use It Is the ProductProduit et UXAI in EducationProduct StrategyAnimeENUnsupervised Learningavr. 2025The Transformer Co-Author Who Says Scaling Hit a WallEntreprise et stratégieIA en entrepriseAI StrategyAnimeENUnsupervised Learningavr. 2025The $177 Billion GPU Bet Only Works If AI Becomes a $1.5 Trillion MarketEntreprise et stratégieAI StrategyCompany BuildingAnimeENUnsupervised Learningmars 2025Open-source models are 5% of enterprise use — and GPT wrappers wonEntreprise et stratégieProduct-Market FitAI StrategyAnimeENUnsupervised Learningmars 2025A Million Tokens per Dollar: Why Noam Shazeer Thinks Inference Compute Is Still UntappedModèles et entraînementReasoning & test-time computeScaling LawsAnimeENUnsupervised Learningmars 2025Mike Schroepfer: AI’s Bottleneck Is Electricity, and One Tanker of Fusion Fuel Could Power the U.S. for a YearInfrastructure et déploiementAI StrategyOpen-weight modelsAnimeENUnsupervised Learningfévr. 2025The Cheaper Intelligence Gets, the More of It We'll BuyAgentsIngénierie d’agentsReinforcement LearningAnimeENUnsupervised Learningjanv. 2025A Million Tokens Cannot Write a Good JokeÉvaluations et observabilitéÉvaluationsBenchmarksAnimeENUnsupervised Learningjanv. 2025A 50,000-GPU cap and a 1,700-chip loopholeInfrastructure et déploiementAI RegulationGeopoliticsAnimeEN
20249 épisodes
Unsupervised Learningdéc. 2024o1 Is a New Generation — and the Day AGI Arrives Will Feel BoringModèles et entraînementReasoning & test-time computeIngénierie d’agentsAnimeENUnsupervised Learningdéc. 2024There will not be one model. Fireworks is betting on hundreds of small expertsInfrastructure et déploiementInferenceModel routing & gatewaysAnimeENUnsupervised Learningdéc. 2024Noam Brown: There's No Hard Wall—Just Eight Orders of Magnitude of HeadroomModèles et entraînementReasoning & test-time computeAGIAnimeENUnsupervised Learningnov. 2024He bet company equity that Transformer would last — then drew an AI map for 12,000 companiesEntreprise et stratégieTransformers & attentionPretraining & midtrainingAnimeENUnsupervised Learningoct. 2024DeepL Beat Google at Translation by Refusing to Be a General ModelEntreprise et stratégieData quality & curationÉvaluationsAnimeENUnsupervised Learningoct. 2024o1 Is Slow and Hard to Use. Percy Liang Still Calls It a Turning PointModèles et entraînementReasoning & test-time computeAI researchAnimeENUnsupervised Learningjuin 2024Scott Belsky: Models Are Commoditizing. Controls, Data, and Taste Are the MoatProduit et UXProduct StrategyAI interface designAnimeENUnsupervised Learningmars 2024A Broken Belt Buckle Turned Gucci's Call Center Into a Sales FloorEntreprise et stratégieIA en entrepriseAI StrategyAnimeENUnsupervised Learningjanv. 2024OpenAI's First DevRel Writes 90% of His Code with ChatGPT — and Says Web Agents Still Need YearsEntreprise et stratégieAI StrategyPrompting & in-context learningAnimeEN
Date inconnue35 épisodes
Unsupervised LearningSuno’s CEO Says the Future of Music Is a Conversation, Not a PromptMultimodal et médiasGenerative mediaMultimodal AIAnimeENUnsupervised LearningPinecone's Vector Database Bet: Scaling Through the ChatGPT TsunamiInfrastructure et déploiementVector searchGénération augmentée par récupérationAnimeENUnsupervised LearningPeter Deng on Why AI Pricing, Evals and Voice Are Still MisunderstoodEntreprise et stratégieProduct StrategyPricing StrategyAnimeENUnsupervised LearningPerplexity's Bet: Search Becomes Answers, and Answers Become a BusinessRAG et connaissancesGénération augmentée par récupérationSearch & retrievalAnimeENUnsupervised LearningCoding Agents Are Breaking Containment: The 2026 AI Stack, the Coding Wars, and the Next FrontierAgents de codeIngénierie d’agentsIngénierie du contexteAnimeENUnsupervised LearningAgents Everywhere: OpenAI's Blueprint for Tool Use, Evals, and Computer-Use InfrastructureAgentsIngénierie d’agentsComputer useAnimeENUnsupervised LearningThe Long Tail, the World Model, and the 1995 Problem: Inside Waymo's Scaling Phase and Robotics' Missing MomentAutonomous DrivingRoboticsMultimodal AIAnimeENUnsupervised LearningSnowflake's AI Stack: Small Models, the Text-to-SQL Iceberg, and Why Governance Comes FirstEntreprise et stratégieInfrastructure et déploiementIA en entrepriseAnimeENUnsupervised LearningCameras Optional: HeyGen's Joshua Xu on Avatars, Orchestration and Building Ahead of the ModelsMultimodal et médiasGenerative mediaIA vocaleAnimeENUnsupervised LearningLegora's CEO on Building Legal AI That Survives the Next Model ReleaseEntreprise et stratégieIA en entrepriseProduct StrategyAnimeENUnsupervised LearningThe Search Engine Under AI Apps: Why Bigger Context Windows Aren't EnoughInfrastructure et déploiementVector searchGénération augmentée par récupérationAnimeENUnsupervised LearningOpenAI’s Platform Bet: Where AI Value Accrues, Why Open Source Trails, and When Superintelligence ArrivesEntreprise et stratégieAGI TimelinesAI StrategyAnimeENUnsupervised LearningFrom RAG to Reasoning: Why the Future of AI Is a System, Not a ModelModèles et entraînementReasoning & test-time computePost-trainingAnimeENUnsupervised LearningServal's Jake Stout on Outbuilding ServiceNow: Mirror the System of Record, Bet on Code, and Ship Before the Incumbent Can Book a MeetingEntreprise et stratégieCompany BuildingIA en entrepriseAnimeENUnsupervised LearningGiving AI a Nervous System: Voice, Touch, and the Post-Keyboard InterfaceMultimodal et médiasIA vocaleMultimodal AIAnimeENUnsupervised LearningOscar Health's AI Playbook: Translate the Bureaucracy First, Then Reinvent CareEntreprise et stratégieAI in healthcareIA en entrepriseAnimeEN
60 sur 79 affichés