筛选结果
101 篇排序
1:08:02EN → 简体中文
Sequoia Capital· 2026年10月13日播客AlphaGo 花了两年，AlphaZero 只用了几个小时：Silver 在超级智能上押了同样的注David Silver 和 Lasse Espeholt 离开 DeepMind，创办了一家二十人的实验室 Ineffable Intelligence。他们想打造一个超级学习者（superlearner），让它从经验中发现科学、数学和语言，而不是靠一轮预训练（pre-training）。
1:04:26EN → 简体中文Sequoia Capital· 2026年10月6日播客100,000 个加速器，每小时都有东西出故障Google AI 基础设施负责人 Amin Vahdat 向红杉资本的 Sonya Huang 解释，为什么 FLOPS 是虚荣指标，为什么 TPU 产品线一分为二，以及为什么他宁愿等电网接入，也不自己建一座吉瓦级的数据中心。
1:05:21EN → 简体中文Sequoia Capital· 2026年9月15日播客从模型到工作流的桥，是一门万亿美元的生意Box 首席执行官 Aaron Levie 认为，这一轮的钱就在前沿模型与真实企业工作之间那片混乱的空档里；编程领域一夜之间的扩散，不会在其他行业重演。
52:27ENSequoia Capital· 2026年9月1日播客After Two Dozen Police Departments Said No, They Built an Anti-Network-Effect AI CompanyA Wisconsin county had 300 gigabytes sitting on a cold case—and Peregrine's founders still refuse to take the credit.
55:19ENSequoia Capital· 2026年8月25日播客Human Click Data Is a Bug: Rebuilding Search for a Web of AgentsParag Agrawal, former Twitter CEO and founder of Parallel Web Systems, argues that agents will query the web a thousand times more than people, and that both search infrastructure and the ad-supported internet have to be rebuilt for that customer.
53:44ENSequoia Capital· 2026年8月18日播客The Biggest Bug in Today’s Models Is That the Weights Never ChangeTuring Award winner Rich Sutton and Khurram Javed founded Oak Lab to build agents that keep learning after they ship — and they call synthetic data a big mistake.
47:22ENSequoia Capital· 2026年8月4日播客From One in a Thousand to 15 Percent: Chai Discovery Treats Drug Design as a Scaling ProblemJoshua Meier and Matthew McPartlon tell Sequoia’s Training Data why antibody design is an engineering problem, not a bespoke science, and why they sell a molecular CAD suite instead of their own drugs.
49:12ENSequoia Capital· 2026年7月29日播客The People Writing the Transformer’s Eulogy Want a Lab That Can Take Vacation Without Themo1’s reasoning lead and a Gemini pre-training lead argue the architecture, not more scale, is now the bottleneck — and that replacing it starts with automating kernel generation.
51:31ENSequoia Capital· 2026年7月21日播客They Refunded Nearly $2 Million. Then Factory’s CEO Explained the Dark FactoryMatan Grinberg on being two years early, why a model-agnostic harness beats lab co-design, and why 90 percent of coding tokens will soon run with the lights off.
48:55EN → 简体中文Sequoia Capital· 2026年7月14日播客Anthropic 的开发者平台一年长出三层，token 也有了分工Katelyn Lesse 和 Angela Jiang 在红杉资本（Sequoia Capital）的《Training Data》播客中讲述，Anthropic 如何从无状态的 Messages API，走到知识、执行和协调三层，同时并不打算自己拥有服务器。
55:19ENSequoia Capital· 2026年7月7日播客The drone is 15% of Zipline: 140 million miles, zero incidentsKeller Rinaudo Cliffton and Eric Watson explain why the aircraft is the small part of an automated logistics system now cheaper than a car.
1:10:15EN → 简体中文Sequoia Capital· 2026年6月30日播客100 倍藏在各层之间：Dylan Patel 谈协同设计，以及一年降本 60 倍SemiAnalysis 创始人认为，让模型、算子和芯片彼此匹配，才能把这里的 2 倍、那里的 2 倍叠加成 100 倍；与此同时，同等质量的推理（inference）成本一年里便宜了约 60 倍。
44:52ENSequoia Capital· 2026年6月24日播客An 80GB Wikipedia Cache Versus a 100GB Model That Holds the InternetEngram’s founders argue that stuffing a company into the context window is the expensive kind of memory—and that baking it into the weights can cut token use by two orders of magnitude.
38:45ENSequoia Capital· 2026年6月16日播客The 25-agent town that became a Fortune 500 decision sandboxOn the day before Valentine's Day in a Stanford simulation, cafe owner Isabella decided to throw a party — no human wrote the plot.
51:09ENSequoia Capital· 2026年6月11日播客Build like the model will eat your harness in a yearLogan Kilpatrick, who leads Google AI Studio and the Gemini API, tells Sequoia's Sonya Huang why agent scaffolding is a 12-month asset, why coding already feels like narrow superintelligence, and why Google is optimizing for outcomes instead of eyeballs.
41:20ENSequoia Capital· 2026年6月10日播客A $4 Million Cabinet and a Trillion-Dollar Year: Jensen Huang Turns the AI Factory Into a Manufacturing StoryOn Sequoia's Training Data stage, NVIDIA's CEO walks a sixty-country audience from ChatGPT's parlor tricks to agentic work, a five-layer industrial cake, and why radiology hiring went up.
40:22ENSequoia Capital· 2026年6月2日播客Building Gets Easier Near AGI. Knowing What to Build Does Not.Listen Labs CEO Alfred Wahlforss on AI interview agents, a 30-million-person audience, and why transcripts beat credit-card data when you need to predict customers.
45:33ENSequoia Capital· 2026年5月26日播客Cursor Ships 1TB of Composer Weights Across Oceans in Under a MinuteFederico Cassano and Fireworks' Dima walk through the globally split RL stack that trained Composer 2 on a sparse trillion-parameter MoE.
38:24ENSequoia Capital· 2026年5月19日播客Serval's Bet: Keep ServiceNow's Primitives, Take Humans Out of the BuildJake Stauch tells Sequoia's Pat Grady how a two-year-old ITSM startup uses codegen agents, boring enterprise controls, and "fewer, better" hiring to rebuild employee support for the AI era.
34:56ENSequoia Capital· 2026年5月13日播客On Suno, 90% of Daily Users Make a SongCEO Mikey Shulman on modeling raw sound instead of music theory, why music is not an LLM-style scale problem, and treating creation itself as the entertainment.
31:26ENSequoia Capital· 2026年4月14日播客ChatGPT Opened 65 Pages to Buy a Shower HeadProfound CEO James Cadwallader argues that the internet's front door is the same, but an agent now walks through it, and marketing has to be rewritten for that visitor.
58:27ENSequoia Capital· 2026年3月24日播客Even a dumber AI can win science if it never leaves the labGinkgo Bioworks CEO Jason Kelly tells Sequoia why autonomous labs, not smarter models, will change how experimental science gets done.
44:19ENSequoia Capital· 2026年3月17日播客The $500-per-Kilogram Line: Why Starcloud Thinks Data Centers Belong in SpaceStarcloud founder Philip Johnston walks through the physics, launch economics, and business model of putting AI compute in orbit.
40:56ENSequoia Capital· 2026年3月10日播客Four of the Top Five Defense Primes Still Run Hardware Tests on PowerPointNominal’s cofounders argue that hardware never got the last twenty years of software infrastructure, and that physical AI will stall until test data lives in one place.加载更多已显示 24 篇，共 101 篇