Results
49 articlesSort
1:34:11ENFrom forty hours in a scanner to one: reconstructing what you see from a shared brain spaceHost Nathan LabenzPaul Scotti explains how MindEye2 maps a new person's fMRI into CLIP with an hour of data, and why a linear layer plus better image generators changed the cost of reading minds.
1:27:01ENThree Million Emails, Ten LLM Calls, Six Models: Shortwave's RAG PlaybookHost Nathan LabenzFirebase founder Andrew Lee walks Nathan Labenz through the retrieval stack that turned a Gmail inbox into a working AI assistant.
1:43:54ENFifteen Examples Can Poison GPT-4: Today's AI Systems Are Exploitable by DefaultHost Nathan LabenzFAR AI founder Adam Gleave walks through cheap jailbreaks in GPT-4's APIs and amateur-level wins against superhuman Go engines, and why robustness is not arriving on its own.
31:28ENBolt Memory onto GPT-4 and Cut Hallucinations by up to 50%Host Sarah Guo, Elad GilPinecone CEO Edo Liberty tells Sarah Guo and Elad Gil how RAG, serverless vectors, and keeping knowledge outside the model make LLMs more reliable.
1:06:39ENOpenAI's First DevRel Writes 90% of His Code with ChatGPT — and Says Web Agents Still Need YearsHost Jacob EffronLogan Kilpatrick on picking GPT-3.5 versus GPT-4, why plugins stalled, and why the public internet is not ready for autonomous agents.
1:22:43ENFive billion minutes of calls: why Dialpad trained its own business LLMHost Nathan LabenzDan O’Connell on real-time speech, verticalizing the stack, and why he and Nathan Labenz disagree about when AI employees show up.
1:21:50ENA 16K Context Window Holds About 350 Rows. That's Why Tables Need an Agent.Host 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.
1:14:48ENDevDay Probably Created Thousands of Startups, Says OpenAI's Logan KilpatrickHost Nathan LabenzThe day after OpenAI's first developer conference, Dev Relations lead Logan Kilpatrick tells Nathan Labenz why thin wrappers were never companies, why agents were deliberately not shipped fully autonomous, and what $2 million custom models are for.
1:32:40ENThe Historian Who Taught GPT-3.5 to Beat GPT-4Host 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.
1:17:09ENGPT-4 as Teacher, GPT-3.5 as Student: A Nine-Tenths Cost Cut and an AI Bundle Nobody Will StartHost Nathan LabenzNathan Labenz reports that chain-of-thought traces from GPT-4 train a cheaper, faster 3.5, then argues for flooring the app accelerator, tapping the brake on GPT-5, and bundling a hundred AI tools for a hundred dollars.
2:32:49ENGPT-4 Failed a Hollow Top Hat. Hollywood's AI Deal Closed Only Half the Loopholes.Host Nathan LabenzWriter Trey Kollmer returns with a strike update, a copyright briefing, and a riddle that still fools the best model.
55:08ENOne Model Inside 38 Products: Paige Bailey on How Google Trains and Ships PaLM 2Host Nathan LabenzAt Google I/O in May 2023, PaLM 2 arrived already running in more than two dozen products — and its lead product manager walks through what that actually took.
52:48ENOpenAI Cut Fine-Tuning 90 Percent, and Founders Crossed Llama 2 Off the ListHost Nathan LabenzNathan Labenz and Erik Torenberg walk through GPT-3.5 Turbo fine-tuning, why self-hosting still loses on total cost, and what is actually working in code, school, and medicine.
1:40:44ENGive the Model Time, Memory, and Tools: Nathan Labenz’s 2023 Field Guide to Using GPT-4Host Nathan LabenzPart two of the AI Scouting Report walks from H100 clusters and RLHF through chain-of-thought, retrieval, agents, and mixture-of-experts.
59:28ENFrom $500,000 to $50,000: MosaicML Made Training Your Own Model a Priced ProductHost Nathan LabenzJonathan Frankle and Abhi Venigalla explain who is training custom LLMs, why extraction beats chat, and how ALiBi stretches a 65k context window.
1:25:35ENTwo and a Half Years From 33.3% to 86.5%: Google Sent Foundation Models to Medical SchoolHost Nathan LabenzVivek Natarajan walks Nathan Labenz through Med-PaLM and Med-PaLM 2: the data-light techniques, the nine-axis clinician evals, and why a high USMLE-style score is the wrong trophy.
42:18ENTen Percent of Search Is Health — and the Doctor Is Googling TooHost Sarah Guo, Elad GilGoogle researcher Karan Singhal on Med-PaLM 2, the real bar for medical language models, and why medicine is a harder alignment lab than chatbots.
37:18ENUse Retrieval for a Wi-Fi Password. Use Experts for a Company's Private Language.Host Sarah Guo, Elad GilGoogle Brain's Kelvin Guu on REALM, FLAN, model surgery, and why personalizing large models is still an unsolved interface problem.
1:21:06ENEveryone Talks About RLHF. Almost Nobody Is Doing It.Host Nathan LabenzHumanloop CEO Raza Habib on why an LLM API is maybe five percent of a working product, and what actually fills the rest.
40:25ENA 6-Billion-Parameter Clone of ChatGPT’s Core TrickHost Sarah Guo, Elad GilDatabricks CTO Matei Zaharia tells Sarah Guo and Elad Gil how Dolly learned instruction-following on fifty times less data than GPT-3 — and why he would rather look facts up than bake them into a trillion-parameter network.
1:56:16ENThey Used Humans as Stand-In AIs. Now 250,000 People Run Elicit.Host Nathan LabenzOught's founders tell Nathan Labenz why open-ended reasoning has to be decomposed, supervised, and grounded—not scaled until it merely looks persuasive.
1:22:36ENStop Chasing AI News: Chip Huyen on What Actually Makes AI Products BetterGuest Chip Huyen · Host Lenny RachitskyAuthor of AI Engineering walks through pre-training, post-training, RL, evals and RAG, then explains what separates companies that get value from AI from those that stall.
23:54ENBeyond GPT-7: The Customization Toolkit, Meta's Open Source Bet and the Return of Consumer AIHost Sarah Guo, Elad GilIn a host-only episode, the No Priors hosts map the six levers for 10x to 100x better AI, explain fine-tuning, RAG and AI feedback, and ask why consumer social is suddenly interesting again.
46:48ENSnowflake's AI Stack: Small Models, the Text-to-SQL Iceberg, and Why Governance Comes FirstHost Jacob EffronSnowflake head of AI Baris Gekken explains how Arctic and Cortex were built, where enterprise AI actually works today, and why precision, governance and evaluation matter more than raw model capability.
Load moreShowing 24 of 49









