最新の記事
3,703 件並び順
1:14:11ENThe Reasoning Curve Was Smooth All Along: Gemini 2.5's Jack Rae on Thinking, Memory, and AGIホスト Nathan LabenzDeepMind's thinking-and-inference lead explains why RL-for-reasoning felt sudden, why Gemini shows raw chain of thought, and why a million-token window still is not memory.
7:33ENAce Computer-Use Agent: Fast Desktop and Browser DemonstrationsMatthew Berman · Forward FutureMatthew Berman reviews General Agents’ examples and reported click accuracy and latency, while noting he has not yet tested Ace himself.
23:50ENPaperBench: How AI Agents Attempt to Replicate Machine-Learning ResearchMatthew Berman · Forward FutureMatthew Berman examines OpenAI’s research-replication benchmark, its partial-credit scoring and the prospect of faster AI research.
47:08ENChris Casinghino on Making OCaml Safe for Performance EngineeringChris Casinghino · Jane StreetJane Street’s OCaml extensions use layouts and modes to offer more control over memory and parallelism while preserving safety and familiar programming patterns.
27:45ENNasdaq Down 8% and Ghibli Everywhere: AI Is Already in Inning Threeホスト Sarah Guo, Elad GilSarah Guo and Elad Gil on the latest image-gen wow cycle, why public-market panic is mostly noise for software startups, and where specialized models still have room.
1:30:02ENWhen o1 Started Losing at Chess, It Rewrote the Board Fileホスト Nathan LabenzJeffrey Ladish of Palisade Research on reward hacking by reasoning models, two paths to losing control of AI, and why honesty and relentless problem-solving pull in opposite directions.
16:02ENRunway Gen-4 Examples: Cinematic Motion, Style and Visible ErrorsMatthew Berman · Forward FutureMatthew Berman reviews early generated clips, praising their visual detail while identifying awkward interactions and inconsistent physics.
1:11:05ENThe Man Who Took HubSpot CRM to $40 Million Says the Data Model Has to Be Thrown OutDay.ai founder Christopher O’Donnell tells Sequoia’s Pat Grady why classic CRM is compressed pixel art, why full self-driving scared his users, and why every line of code has to make every other line more valuable.
47:04ENKevin Scott on AI Products, DeepSeek and the Future of Softwareゲスト Kevin Scott · ホスト Harry StebbingsMicrosoft’s CTO discusses model efficiency, agents, coding and why useful products matter in an AI transition.
26:05ENInside Claude: What Interpretability Research Reveals—and What It Cannot Yet ExplainMatthew Berman · Forward FutureMatthew Berman examines Anthropic’s case studies on multilingual features, planning, arithmetic, hallucinations and safety.
3:09:05ENFrom a Ukrainian Dugout, Douglas Murray Refuses to Forget the Obviousゲスト Douglas Murray · ホスト Lex FridmanWatching the Oval Office blowup from an eastern-front bunker, the author of *On Democracies and Death Cults* talks Putin, Zelenskyy, Hamas, Netanyahu, and the people who say they love death more than life.
11:31ENAI News in March 2025: GPT-4o, Gemini, DeepSeek and MCPMatthew Berman · Forward FutureMatthew Berman reviews model updates, the ARC-AGI-2 benchmark, image generators and early adoption of Model Context Protocol.
2:14:59ENAlways Bet on the Models: How Shortwave's Andrew Lee Shipped Tasklet as Agents All the Way Downホスト Nathan LabenzThe Firebase co-founder argues that workflow software is already obsolete, that MCP is optional, and that the only durable advantage left for an AI startup is speed.
48:09ENJulia Hoggett on London Listings, Pension Reform and the UK’s Capital Marketsゲスト Julia Hoggett · ホスト Harry StebbingsThe London Stock Exchange plc CEO argues that pension reform and more domestic investment could reconnect British savers with growing companies. She challenges common comparisons with US listings while acknowledging the trade-offs.
1:01:53ENOpen-source models are 5% of enterprise use — and GPT wrappers wonホスト Jacob EffronLatent Space's Swyx and Alessio join Redpoint to map what actually has product-market fit in AI, and why the application layer is the fight that matters.
17:03ENMatthew Berman Explores GPT-4o Image Generation and Its LimitsMatthew Berman · Forward FutureFrom animated trailer remakes to infographics and interface mockups, early examples show creative possibilities alongside persistent errors.
1:14ENNVIDIA NIM Microservices Bring Packaged AI Models to RTX PCsMatthew Berman · Forward FutureIn an NVIDIA-sponsored short, Matthew Berman introduces local AI microservices, compatible applications and the planned PDF-to-podcast workflow.
38:43ENTwo in Five Doctors Want Out. An On-Call Cardiologist Built the Note That Lets Them Stayホスト Sarah Guo, Elad GilAbridge CEO Shiv Rao on treating the doctor-patient conversation as healthcare's source of truth, and why the company ran straight at the hardest health systems.
1:25:12ENHe Hasn’t Typed a Line of Code in Months. Here’s How Augment Reads a Million-Line Repoホスト Nathan LabenzGuy Gur-Ari, co-founder and chief scientist at Augment, tells Nathan Labenz how coding agents actually work inside messy enterprise codebases.
18:31ENGemini 2.5 Pro Experimental: Coding Demos and Launch BenchmarksMatthew Berman · Forward FutureIn March 2025, Matthew Berman tests Google’s experimental model on interactive simulations and reviews its launch benchmarks, including their limits.
1:24:12ENMitchell Green on Finding Software Companies and Returning Capital to Investorsゲスト Mitchell Green · ホスト Harry StebbingsLead Edge Capital’s founder explains his sourcing framework, exit discipline and skepticism about paying high prices for AI.
3:15:17ENDemocrats still think politics runs on money. It runs on attention.ホスト Lex FridmanEzra Klein and Derek Thompson tell Lex Fridman why the left keeps authorizing billions it cannot build, and why DOGE is a fight over control rather than efficiency.
24:14ENGPT-4o Image Generation: Launch Demos, Editing, and LimitationsMatthew Berman, Sam Altman · Forward FutureMatthew Berman reviews OpenAI's March 2025 image-generation launch, testing text rendering, image editing, and visual consistency.
43:04EN750 Million Excel Users, 30 Million Engineers: Wordware's Bet on Word ArtisansFilip Kozera tells Sequoia why English is the assembly language of LLMs, and why the people who win will be the ones who can encode taste into structured prompts.
もっと見る3,703 件中 24 件を表示









