최신 글
4,474개정렬
1:02:01ENJoost de Valk on Building Yoast, SEO Fundamentals and the AI Content Shift게스트 Joost de Valk · 진행자 Harry StebbingsThe Yoast founder explains how a free WordPress plugin became a business, what founders get wrong about search, and how AI may change content and software.
8:02ENMeta’s Segment Anything: Promptable Image Masks and the SA-1B DatasetMatthew Berman · Forward FutureIn April 2023, Matthew Berman reviews Meta’s Segment Anything project and demonstrates selecting image regions with clicks, boxes and automatic mask generation. The model segments pixels; it does not by itself name every object or guarantee accurate results on every image.
11:38ENHow Generative Agents Populate a Simulated TownMatthew Berman · Forward FutureMatthew Berman explores a Stanford and Google research prototype in which language-model agents remember, plan and interact in a small virtual town.
1:09:54ENA $100 Doctor Visit and a $1 AI: GPT-4 Was Already Unbundling Work진행자 Nathan LabenzBefore GPT-4 launched, Nathan Labenz spent two months making the model play a dozen paid professions, then argued with Erik Torenberg about what that does to jobs, regulation, and white-collar status.
3:46:39ENHe Hated School Math Until a Professor Opened the Book to Quarks게스트 Edward Frenkel · 진행자 Lex FridmanEdward Frenkel talks with Lex Fridman about Soviet childhood, the Langlands program, Gödel, AI, and why every hard objective theory still ends with an observer.
12:59ENAuto-GPT, Memory and Model Orchestration: Early Agent ExperimentsA 2023 snapshot of iterative prompts, tool use, persistent context and AI safety questions.
1:10:16ENRob Lacher on Building a Venture Firm and Europe’s Entrepreneur Networks게스트 Rob Lacher · 진행자 Harry StebbingsThe Visionaries Club founder discusses hiring, fund concentration, follow-on reserves and the role of family businesses in European technology.
40:25ENA 6-Billion-Parameter Clone of ChatGPT’s Core Trick진행자 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.
6:24ENBloombergGPT: Mixing Financial and General Data for a Language ModelMatthew Berman · Forward FutureMatthew Berman examines the 2023 finance-focused research model, its training corpus, evaluation and decision not to release weights.
1:56:16ENThey Used Humans as Stand-In AIs. Now 250,000 People Run Elicit.진행자 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.
0:29ENBill Ackman Proposes Taxing Some Borrowing Against Assets게스트 Bill Ackman · 진행자 Harry StebbingsA 2023 policy argument about stock and real-estate borrowing.
7:16ENSetting Up Alpaca Turbo for Local Language Models: A 2023 TutorialMatthew Berman · Forward FutureMatthew Berman demonstrates an early ChatGPT-style interface for local Alpaca models, from the Conda setup to a short test on a Mac.
1:33:36ENMike Salguero on Bootstrapping ButcherBox and Subscription Economics게스트 Mike Salguero · 진행자 Harry StebbingsThe ButcherBox founder explains customer acquisition, per-box margins and the trade-offs behind building without venture capital.
1:33:49ENSafe Only Because It's Weak: A GPT-4 Red-Teamer and Two Founders Debate the Pause진행자 Nathan LabenzNathan Labenz, who spent six months trying to make GPT-4 misbehave, wants labs to stop training anything bigger; Chroma's Anton Troynikov and Lindy's Flo Crivello argue a six-month freeze buys little and may select for the least careful labs.
3:34:58ENAn 18-year-old bought a one-way ticket to the Amazon게스트 Paul Rosolie · 진행자 Lex FridmanPaul Rosolie had no degree, no funding, and no expedition. Seventeen years later he is still there, protecting a library of life we have not finished reading.
7:53ENHuggingGPT: Using a Language Model to Coordinate AI ToolsMatthew Berman · Forward FutureMatthew Berman explains a research system that routes multimodal tasks to specialist models using a language model as controller.
5:08ENChatting with a PDF in Colab Using LangChain and a Vector IndexMatthew Berman · Forward FutureMatthew Berman demonstrates a 2023 PDF question-answering notebook; its retrieval workflow does not fine-tune model weights and its revenue example is incorrect.
1:23:23ENTobi Lütke on Learning, Creative Constraints and Leading Shopify게스트 Tobi Lütke · 진행자 Harry StebbingsThe Shopify CEO explains how he updates his mental models, why constraints can help teams create and how he weighs remote work against proximity.
7:48ENUsing ChatGPT - Genie AI in VS Code to Explain, Refactor and Test Rails CodeMatthew Berman · Forward FutureMatthew Berman demonstrates a 2023 VS Code extension on a Ruby on Rails project and examines the limits of its code context.
1:19:13ENJackie Reses on Hiring, Banking Infrastructure and the Future of Fintech게스트 Jackie Reses · 진행자 Harry StebbingsThe Lead Bank CEO explains her approach to talent and negotiation, and offers a 2023 view of fintech, crypto and regulation.
41:40ENEvery 10 Milliseconds Cost Copilot 1% of Completions진행자 Sarah Guo, Elad GilGitHub Copilot's chief architect, Alex Graveley, walks through how a leftover code model with a sub-10% test pass rate became a product developers would not quit — and what he is building next at Minion AI.
11:28ENCan AI Development Be Paused? The 2023 Open Letter and Its Trade-OffsMatthew Berman, Sam Altman · Forward FutureMatthew Berman examines the call to pause training systems more powerful than GPT-4, the concerns behind it and the difficulties of coordination.
3:17:51ENWhoops is the sound of getting less wrong: Yudkowsky, two weeks after GPT-4게스트 Eliezer Yudkowsky · 진행자 Lex FridmanTwo weeks after GPT-4 launched, Eliezer Yudkowsky sat down with Lex Fridman and started by admitting the transformer stack had already gone further than he expected.
1:29:31EN130,000 Teleop Demos: Google's GPT Playbook for Robots진행자 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.
더 보기4,474개 중 24개 표시









