絞り込み結果
20 件並び順
1:04:51EN125 million cells took a decade. The next billion takes monthsホスト 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.
1:32:38ENPolish Is 1% of Frontier Training Data, So Poland Built Its Own Modelホスト 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.
1:01:49ENChatGPT's Co-Creator Left to Put Nature in the Reward Loopホスト Nathan LabenzLiam Fedus and Ekin Dogus Cubuk on why language models still fail at physics, why literature data is too noisy to train on, and why Periodic Labs is aiming at a 200K superconductor.
1:22:07ENEvery payment is a frame in a movie. Stripe’s foundation model took card-testing detection from 59% to 97%ホスト Nathan LabenzEmily Sands, Stripe’s head of data and AI, explains how a domain-specific payments model, shared embeddings, and mixed rules turn a $1.4 trillion network into compounding advantage.
1:55:02ENA hundred experiments a day, and the valley of death in materialsホスト Nathan LabenzRadical AI’s co-founders tell Nathan Labenz how a closed loop of frontier models and a self-driving lab is meant to collapse a forty-year materials cycle into less than a decade.
1:42:06ENEvery Frontier Lab Is Spending a Billion a Year on Data. Here Is Where It Goesホスト Nathan LabenzLabelbox CEO Manu Sharma on the $15 billion Scale deal, RL “gyms,” $250,000 trainers, and why post-training now eats the budget that used to belong to pre-training.
1:43:54ENA Year to Grow a Brain Organoid, an Hour to Simulate Itホスト Nathan LabenzStanford postdoc Siyu He tells Nathan Labenz how Squidiff and CORAL move single-cell and spatial experiments into the computer — and where the models still cannot go.
1:40:22ENTwenty-atom crystals taught an AI the physics of a heartbeatホスト Nathan LabenzOrbital Materials' Jonathan Godwin and Tim Duignan tell Nathan Labenz how a neural network trained on tiny crystals simulated a potassium ion channel on a single 2017 GPU.
1:27:33ENA 70B Model Booked Dinner 20% of the Time. MultiOn Took It to 95% in a Dayホスト Nathan LabenzMultiOn CEO Div Garg returns to The Cognitive Revolution to recap 18 months of agent hype, the AgentQ stack, and why 2025 might still be the year of vertical agents.
1:49:41ENA Team of Fifteen Matched Meta’s Post-Training on Meta’s Own Baseホスト Nathan LabenzNathan Lambert walks through Tulu 3: the three-stage recipe, the 1 percent versus 14 percent lesson, and why open post-training now lives or dies on data and evals.
1:13:23ENAn AI Found a GABA-Cancer Link. The Next Step Is Talking to the Cells.ホスト Nathan LabenzMichael Levin and Leo Pio Lopez on a multilayer embedding that predicted melanoma, a cancer with no mutations, and why biology's real interface is cognitive.
1:16:11ENScooping Food Into Boxes Is 70 Percent of Plant Laborホスト Nathan LabenzChef Robotics CEO Rajat Bhageria on 20 million plated meals, a foundation model for food, and why the kitchen may leave the house.
1:01:39ENGPT-5 Wants 100 Trillion Tokens. Does Earth Have Enough?ホスト Nathan LabenzTraining GPT-3 took about a trillion tokens; GPT-4 took about ten trillion. The next jump is a data question as much as a compute one.
1:21:50ENA 16K Context Window Holds About 350 Rows. That's Why Tables Need an Agent.ホスト Nathan LabenzGretel AI founder Alex Watson tells Nathan Labenz why most company data is still a spreadsheet, why GPT-4 is the wrong tool for generating it, and how differential privacy lets a model learn a disease without learning a patient.
1:32:40ENThe Historian Who Taught GPT-3.5 to Beat GPT-4ホスト Nathan LabenzMark Humphries on a kilometer of unread paper, fur-trade jargon that embeddings cannot see, and why cheap processing will force history to be rewritten.
2:15:11ENTwenty Gibberish Tokens Jailbroke ChatGPT and Bardホスト Nathan LabenzCarnegie Mellon’s Zico Kolter, Andy Zou, and Asher Trockman walk through a universal jailbreak that transfers from open models to commercial APIs, and a training-free way to steal some of pretraining’s benefit.
59:28ENFrom $500,000 to $50,000: MosaicML Made Training Your Own Model a Priced Productホスト Nathan LabenzJonathan Frankle and Abhi Venigalla explain who is training custom LLMs, why extraction beats chat, and how ALiBi stretches a 65k context window.
2:00:19EN1.5 Million GPT-4 Fairy Tales Let a 33-Million-Parameter Model Out-Reason GPT-2ホスト Nathan LabenzTry this fill-in: Lily likes cats and dogs. She asked her mom for a dog. Mom said no, so she asked for a…
1:29:31EN130,000 Teleop Demos: Google's GPT Playbook for Robotsホスト Nathan LabenzKeerthana Gopalakrishnan of Google Robotics tells Nathan Labenz how RT-1, SayCan, and PaLM-E turn language-model scaling into kitchen-ready mobile manipulators.
1:03:59ENBuilding a World Model of Life: Biohub's Alex Rives on Why Better Measurement Must Come Before Virtual Cellsホスト Nathan LabenzBiohub's head of science explains why today's virtual cell models fall short, how a $1.8 billion data push aims to fix it, and what a five-year information-level model of the cell could unlock.








