Derniers articles
4 470 articlesTrier
12:20ENTesting UltraLM 13B: A Small Model’s Strengths and Reasoning ErrorsMatthew Berman · Forward FutureA 2023 local test finds useful writing and simple coding alongside failures in games, arithmetic and logic.
10:44ENCan AI Help Humans Align More Powerful AI?Matthew Berman · Forward FutureMatthew Berman examines OpenAI’s 2023 Superalignment proposal, its plans for AI-assisted oversight, and the unanswered question of who evaluates the evaluators.
2:28:37ENThe Marshals Came for an Underground City. RFK Jr. Still Wants a President Who Can Say NoInvité : Robert F. Kennedy Jr · Anime : Lex FridmanOn Independence Day 2023, Lex Fridman sat with Robert F. Kennedy Jr. for two and a half hours on Cuba, Ukraine, the CIA, pharma, and why the shopping cart still has to go back.
1:33:56ENThe AI Scouting Report: 90 Minutes From GPT-4's Exam Scores to Whether Models UnderstandAnime : Nathan LabenzNathan Labenz spends a solo episode walking through what general-purpose models can do, how they are trained, how capabilities unlock, and whether they understand anything at all.
9:01ENInstalling and Testing DragGAN Locally: Matthew Berman’s 2023 WalkthroughMatthew Berman · Forward FutureA historical walkthrough of point-based image editing, dependency workarounds and the limits of an early DragGAN release.
1:40:07ENThirty thousand voltage updates a second: inside Skydio’s autonomy stackAnime : Nathan LabenzHayk Martiros walks Nathan Labenz through a two-pound drone that sees in every direction, runs more than ten neural nets on board, and will not be weaponized.
2:20ENSean on Hospitality and Office Life at Jane StreetSean · Jane StreetA member of Jane Street's London hospitality team describes fresh food, teamwork and the office events that bring colleagues together.
16:39ENPhi-1 and the Value of Textbook-Quality Training DataA 2023 analysis of a small coding model, synthetic data, repeated training and uncertain AI timelines.
7:05ENAider’s Early Coding Assistant: Snake and an Existing Rails ProjectMatthew Berman · Forward FutureMatthew Berman demonstrates the terminal tool’s 2023 workflow, from generating and iterating on a game to proposing a dependency change in a repository.
1:47:05ENEdit Meaning, Not Tokens: Linus Lee on Latent-Space InterfacesAnime : Nathan LabenzThe Notion AI product lead walks Nathan Labenz through the interfaces, model autopsies, and agency questions behind his research prototypes.
9:33ENTesting SDXL 0.9: Image Quality, Clipdrop and a Midjourney ComparisonMatthew Berman · Forward FutureIn June 2023, Matthew Berman tries Stability AI’s image model, compares selected prompts with Midjourney and demonstrates Clipdrop’s lighting tool.
2:09ENDan on Trading Desk Operations at Jane StreetDan · Jane StreetA London trading desk operations team member explains the variety of the role, what interviews seek, and why asking questions matters.
2:21ENVarun on Software Engineering at Jane StreetVarun · Jane StreetA London-based developer describes useful technical foundations, OCaml training and team matching at Jane Street.
54:00ENDouwe Kiela on Data, Retrieval and Enterprise AIInvité : Douwe Kiela · Anime : Harry StebbingsThe Contextual AI CEO discusses language model trade-offs, evaluation, open models and the hurdles to enterprise use.
3:08:46ENWho owns the off switch? George Hotz on tinygrad, open-source AI, and GodInvité : George Hotz · Anime : Lex FridmanOn his third Lex Fridman podcast, the comma.ai and tiny corp founder argues that compute, intelligence, and the off switch should live in people's houses, not in a trusted few.
51:22ENThree IPOs Later, Frank Slootman Still Thinks the Job Is UrgencyAnime : Sarah Guo, Elad GilSnowflake's CEO tells No Priors how he picked the right elevator, turned a 15-person backup startup into an EMC acquisition, and why a data lake is a landfill you should not train on.
18:08ENSix Language Models Compared on an Informal 2023 Prompt SetMatthew Berman · Forward FutureMatthew Berman tests coding, writing and reasoning with H2O GPT, finding a tie under a small and subjective rubric.
1:32:22ENA Million Bytes and 256 Symbols: Meta's MEGABYTE Wants to Throw the Tokenizer AwayAnime : Nathan LabenzLili Yu of Meta AI explains a multiscale transformer that predicts million-byte sequences, skips tokenization, and aims to model any file as raw bytes.
1:04:52ENJennifer Hyman on Building Rent the Runway and Managing for the Long TermInvité : Jennifer Hyman · Anime : Harry StebbingsThe founder discusses hiring, rental economics, paid marketing, board leadership and the work of turning a fashion subscription into a durable business.
5:02ENMatthew Berman on Publishing Videos About Unrestricted AI ModelsMatthew Berman · Forward FutureAfter viewers raise concerns, Berman explains how he weighs freedom of information, research and the risk of misuse.
1:52:46ENCharge Neural Nets by Connection Length, and the Black Box Splits Into Modules You Can SeeAnime : Nathan LabenzMIT physicist Ziming Liu walks Nathan Labenz through BIMT, a brain-inspired training trick that sparsifies networks until you can read them by looking.
59:36ENPascal Pfeiffer on Retrieval, Open Models, Kaggle and Fine-Tuning with H2O LLM StudioMatthew Berman, Pascal Pfeiffer · Forward FutureIn a 2023 interview, H2O.ai’s Pascal Pfeiffer explains document retrieval and parameter-efficient fine-tuning, reflects on Kaggle competitions and demonstrates LLM Studio. He and Matthew Berman also consider coding assistants and unverified rumors about model architecture.
27:25ENConvolution: Adding Probability DistributionsGrant Sanderson · 3Blue1BrownTwo visualizations connect weighted dice, continuous densities and the central limit theorem.
1:52:28ENAkin Babayigit on Building Tripledot, Mobile Games and Questioning Startup AdviceInvité : Akin Babayigit · Anime : Harry StebbingsThe Tripledot co-founder discusses execution, hiring, gaming economics and why founders should treat universal startup rules with care.
Charger plus24 sur 4 470 affichés









