Recently updated channels

Pick one to see only that channelBrowse all channels →

Results

217 articles
Sort
2:51:18EN
Zhang Xiaojun Podcast· Oct 15, 2026Podcast · 2 speakersLindon Gao on Building Dyna Robotics and Finding a Path Through the FogGuest Lindon Gao · Host 张小珺Dyna Robotics CEO Lindon Gao explains why his company builds intelligence through real workflows, how hardware and research must evolve together, and how a founder can find clarity by letting go of the need to be right.RoboticsProduct StrategyModels & training+2
5 slides10:36EN
MBMatthew Berman· Oct 13, 2026Meta Muse Wonder Tested: Strong Benchmarks, Mixed Demos and Quota FrictionMatthew Berman · Forward FutureMatthew Berman tests Meta’s models on browser demos and reviews their published benchmark scores, availability and cost.Models & trainingBenchmarksAI cost & performance+1
1:08:02EN
Sequoia Capital· Oct 13, 2026PodcastAlphaGo Took Two Years. AlphaZero Took Hours. Silver Is Making the Same Bet on SuperintelligenceDavid Silver and Lasse Espeholt left DeepMind to build Ineffable Intelligence, a twenty-person lab that wants a superlearner to discover science, math, and language from experience rather than from a pre-training run.Models & trainingReinforcement LearningRL environments & simulators+3
56:31EN
Unsupervised Learning: With Jacob Effron· Oct 6, 2026PodcastPost-Training Wins Inference: Why Owning the Weights Beats Betting on One LabHost Jacob EffronApplied Compute CEO Yash Patil argues that control, private evals, and post-training on the biggest workloads matter more than fearing that frontier labs will turn malicious.Models & trainingPost-trainingReinforcement Learning+3
5 slides16:57EN
MBMatthew Berman· Sep 22, 2026Grok 4.7 Compared: Coding, Knowledge Work and Cost per TaskMatthew Berman · Forward FutureMatthew Berman examines reported benchmarks and prices while noting missing competitors and limited hands-on testing.BenchmarksAI cost & performanceCoding agents+1
38:14EN
No Priors: AI, Machine Learning, Tech, & Startups· Sep 18, 2026Podcast · 2 speakersDiffusion Will Win Inference: Stefano Ermon Bets Parallel Tokens Beat Custom SiliconHost Sarah Guo, Elad GilThe Stanford diffusion pioneer tells No Priors why a 50-person lab is training language models that generate many tokens at once — and why that maps to GPUs the way transformers once mapped to training.Models & trainingInferenceAI cost & performance+3
1:56:40EN
Cognitive Revolution "How AI Changes Everything"· Aug 8, 2026PodcastThe Periodic Table Was Already in the WeightsHost Nathan LabenzGoodfire CTO Dan Balsam tells Nathan Labenz how concept manifolds, predictive data debugging, and a $1,000-a-month research agent are turning interpretability into a product.Models & trainingModel InterpretabilityAgent engineering+3
47:22EN
Sequoia Capital· Aug 4, 2026PodcastFrom One in a Thousand to 15 Percent: Chai Discovery Treats Drug Design as a Scaling ProblemJoshua Meier and Matthew McPartlon tell Sequoia’s Training Data why antibody design is an engineering problem, not a bespoke science, and why they sell a molecular CAD suite instead of their own drugs.Models & trainingAI for scienceScaling Laws+3
49:12EN
Sequoia Capital· Jul 29, 2026PodcastThe People Writing the Transformer’s Eulogy Want a Lab That Can Take Vacation Without Themo1’s reasoning lead and a Gemini pre-training lead argue the architecture, not more scale, is now the bottleneck — and that replacing it starts with automating kernel generation.Models & trainingAI researchPretraining & midtraining+3
2:59:03EN
Zhang Xiaojun Podcast· Jul 28, 2026Podcast · 2 speakersYou Kaichao on How vLLM Grew from Open-Source Project to AI Infrastructure CompanyGuest You Kaichao · Host 张小珺Inferact co-founder You Kaichao traces vLLM’s path from Berkeley research to a company, then explains why model, inference, hardware and agent design increasingly have to move together.Company BuildingAI startups & small teamsAI research+2
3:46:12EN
Zhang Xiaojun Podcast· Jul 16, 2026Podcast · 2 speakersKay Ke on Physical Intelligence, Robot Learning, and the Future of RoboticsGuest Kay Ke · Host 张小珺Physical Intelligence researcher Kay Ke traces the field's research lineages and explains Pi's progression from π₀ to π*₀.₆. She also reflects on education, fiction, ambition, and what it means to build robots that work beyond a demo.RoboticsReinforcement LearningAI research+2
2:06:41EN
Cognitive Revolution "How AI Changes Everything"· Jul 9, 2026PodcastZero the J-Space and a Model's Hidden Goal Lights Up on Token OneHost Nathan LabenzA weekly Cognitive Revolution cut of Anthropic's global-workspace paper, AI superforecasters, open video world models, and a chip that treats inference as data movement.Models & trainingModel InterpretabilitySafety & governance+3
50:57EN
Unsupervised Learning: With Jacob Effron· Jul 9, 2026PodcastA Trillion Dollars of GPUs Will Be Worth $100 Billion in Five Years, Says LSTM's Jürgen SchmidhuberHost Jacob EffronThe AI pioneer tells Unsupervised Learning's Jacob Effron why he is wildly optimistic about the technology and deeply pessimistic about the model companies buying the chips.Models & trainingAGIAI research+3
1:47:38EN
Cognitive Revolution "How AI Changes Everything"· Jul 4, 2026Podcast19 Neurons Can Drive a Car. Liquid AI Wants Foundation Models on Every ChipHost Nathan LabenzRamin Hasani traces liquid neural networks from a worm's 300-cell brain to a 600-megabyte model that will talk from inside a Mercedes, and explains why architecture search now beats architecture avengers.Models & trainingLocal inferenceScaling Laws+3
2:59:44EN
Cognitive Revolution "How AI Changes Everything"· Jun 3, 2026PodcastTwo Endangered Languages Broke Transformers. Nested Learning Gave Models a Sleep Button.Host Nathan LabenzGoogle researcher Ali Behrouz tells Nathan Labenz how stacking update frequencies — not just layers — and an offline sleep phase could give language models genuine continual learning.Models & trainingContinual learningTransformers & attention+3
1:13:33EN
Unsupervised Learning: With Jacob Effron· Jun 3, 2026PodcastA Trillion Tokens Later, the Concept Finally Lands: Łukasz Kaiser on What Transformers Still MissHost Jacob EffronThe Attention Is All You Need co-author tells Unsupervised Learning that transformers will probably keep winning — and that they still do not learn the way people do.Models & trainingAI researchCoding agents+3
59:41EN
Unsupervised Learning: With Jacob Effron· May 22, 2026PodcastThe GPT moment for video hasn’t arrived, but AGI may already be hereHost Jacob EffronA day after Google I/O 2026, Gemini co-lead Oriol Vinyals argues that world models, file-system memory, and infinite-complexity RL environments will define the next era of frontier models.Models & trainingWorld modelsReinforcement Learning+3
1:21:56EN
Unsupervised Learning: With Jacob Effron· May 15, 2026PodcastYann LeCun Left Meta Because LLMs Will Not Get You to IntelligenceHost Jacob EffronThe Turing Award winner tells Jacob Effron why he spun AMI Labs out of Meta, why JEPA-style world models beat pixel generators, and why he thinks Hinton changed his mind in 2023.Models & trainingWorld modelsAI research+3
1:48:43EN
Cognitive Revolution "How AI Changes Everything"· May 1, 2026PodcastThe reward jumped from 3 to 180. Every Hacker News title was a Google layoff.Host Nathan LabenzCoreWeave's Kyle Corbitt walks Nathan Labenz through GRPO, rubrics, RL environments, and why most reward hacks are easy to catch once you know to look.Models & trainingReinforcement LearningPost-training+3
58:46EN
Unsupervised Learning: With Jacob Effron· Apr 9, 2026PodcastTwo Dates from OpenAI's Chief Scientist: Intern by September, Automated Researcher by 2028Host Jacob EffronJakub Pachocki tells Jacob Effron why those timelines still hold, why continual learning is not a detour, and why showing chain of thought in the product would train the monitor away.Models & trainingContinual learningReinforcement Learning+3
1:49:53EN
Cognitive Revolution "How AI Changes Everything"· Mar 5, 2026PodcastDon't Fight Backprop: Goodfire's Bet on Designing What Models LearnHost Nathan LabenzDan Balsam and Tom McGrath return to The Cognitive Revolution with a $150 million Series B, a new research pillar called intentional design, and a rule for anyone who wants to steer training: stop trying to out-muscle gradient descent.Models & trainingModel InterpretabilitySafety & governance+3
1:34:29EN
Cognitive Revolution "How AI Changes Everything"· Feb 18, 2026PodcastThree IMO Golds Missed the Same Problem. Harmonic Wants Math to CompileHost Nathan LabenzVlad Tenev and Tudor Achim built Aristotle to emit Lean proofs a kernel can check—and they argue that is how mathematical superintelligence stays trustworthy.Models & trainingReasoning & test-time computeAI research+3
1:04:51EN
Cognitive Revolution "How AI Changes Everything"· Feb 1, 2026Podcast125 million cells took a decade. The next billion takes monthsHost Nathan LabenzMark Zuckerberg and Priscilla Chan on pairing a frontier biology lab with a frontier AI lab so models train on data that was designed for them.Models & trainingAI for scienceData quality & curation+3
4:25:13EN
Lex Fridman· Jan 31, 2026PodcastThe Architecture Barely Changed. Data, Cost, and Open Weights Did.Host Lex FridmanNathan Lambert and Sebastian Raschka walk Lex Fridman through a year in which training methods and datasets moved more than model design, and 2026 looks like a fight over inference cost, agents, and who owns truly open models.Models & trainingScaling LawsOpen-weight models+3
Load moreShowing 24 of 217