筛选结果
9 篇排序
1:04:14ENThe Man Who Helped Stand Up Colossus in 120 Days Is Betting on an AI That Lives in Your House主持 Jacob EffronxAI co-founder Igor Babuschkin left to start River AI, arguing coding agents have crossed a threshold, closed labs are getting squeezed, and the next prize is personal, local intelligence.
1:47:38EN19 Neurons Can Drive a Car. Liquid AI Wants Foundation Models on Every Chip主持 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.
1:32:38ENPolish Is 1% of Frontier Training Data, So Poland Built Its Own Model主持 Nathan LabenzMarek Kozlowski of Poland's National Information Processing Institute explains how PLLuM uses language adaptation, organic instructions, and on-prem small models to keep a mid-size country in control of its AI.
2:24:18EN → 简体中文MMLU 在训练后期才脱离随机水平,loss 只是一个警报主持 Nathan LabenzZyphra 的训练负责人讲解了 Zamba 2 的混合架构、真正重要的评测,以及小型实验室仍需重新摸索的那些技术诀窍。
1:16:11ENScooping Food Into Boxes Is 70 Percent of Plant Labor主持 Nathan LabenzChef Robotics CEO Rajat Bhageria on 20 million plated meals, a foundation model for food, and why the kitchen may leave the house.
34:09ENFive Independent Teams Found the Same 10-to-1 Mix That Beats Transformers主持 Sarah Guo, Elad GilThe authors of S4 and Mamba argue that mostly state-space layers plus a dash of attention is the architecture that actually works—and they are shipping it first as real-time speech.
29:09ENThe Hyperscalers Are Spending $200 Billion — and the Next Bottleneck Is a Power Plant主持 Sarah Guo, Elad GilSarah Guo and Elad Gil walk through music models, on-device LLMs, long context, and why AI capex now sits beside oil and broadband.
38:33ENSpeed Is the Thing You Can Buy: Baseten’s Bet on Inference主持 Sarah Guo, Elad GilTuhin Srivastava tells Sarah Guo and Elad Gil why a six-person chatbot company running a billion tokens a day should not be building its own serving stack.
1:18:48ENData Movement, Not Math: How to Make Deep Learning Fit on Robots and Phones主持 Lex FridmanMIT professor Vivienne Sze explains why moving data dominates the energy cost of neural networks and how co-designing hardware and algorithms can cut it by orders of magnitude.









