最近更新的频道

点选即可只看该频道浏览全部频道 →

筛选结果

49 篇
排序
1:29:02EN
Cognitive Revolution "How AI Changes Everything"· 2026年7月1日播客From 50 Designs a Day to 1,500: How AI Changed What Jaguar Land Rover Can Explore主持 Nathan LabenzNeural Concept’s Thomas von Tschammer on specialist physics models, engineering copilots, and why Formula 1 already runs thousands of designs overnight.企业与战略微调Agent orchestration+3
44:52EN
Sequoia 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.RAG 与知识库Continual learning微调+3
1:48:43EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2026年5月1日播客奖励从 3 涨到 180,Hacker News 上的标题全是谷歌裁员主持 Nathan LabenzCoreWeave 的 Kyle Corbitt 带 Nathan Labenz 讲 GRPO、评分标准(rubric)、强化学习环境,以及为什么大多数奖励作弊(reward hacking)只要知道去找,就很容易发现。模型与训练Reinforcement LearningPost-training+3
2:25:10EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2026年1月22日播客微调没有死,却把一个乐于助人的模型变得邪恶了主持 Nathan LabenzNathan Labenz 谈为什么大多数工作改由提示词完成而不再微调,涌现式错位(emergent misalignment)到底揭示了什么,以及他如何看待工作、全民基本收入(UBI)和通用人工智能(AGI)。模型与训练微调安全与治理+3
1:32:38EN
Cognitive Revolution "How AI Changes Everything"· 2025年12月6日播客Polish Is 1% of Frontier Training Data, So Poland Built Its Own Model主持 Nathan LabenzMarek Kozlowski of Poland's National Information Processing Institute explains how PLLuM uses language adaptation, organic instructions, and on-prem small models to keep a mid-size country in control of its AI.企业与战略Open-weight modelsData quality & curation+3
1:55:12EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2025年11月8日播客28,000 份文献、一个 30 亿参数的模型、22 个人:天主教 AI 的技术账本主持 Nathan LabenzLongbeard 首席执行官 Matthew Harvey Sanders 在 The Cognitive Revolution 节目中讲述,一支使命驱动的团队如何让模型与天主教训导对齐,以及为什么微调还不够用。模型与训练检索增强生成微调+3
50:00EN
Sequoia Capital· 2025年9月2日播客Crosby Bills by the Document, Not the Hour, and Turns Contracts Around in Under an HourThe founders of an AI-native law firm explain why they became lawyers with engineers at the next desk instead of shipping legal software, and how deal velocity became the product.企业与战略微调上下文工程+3
38:10EN
Cognitive Revolution "How AI Changes Everything"· 2025年7月16日播客Plus Five for Format, Plus Ten if It Compiles主持 Nathan LabenzPredibase CEO Dev Rishi explains how reinforcement fine-tuning, production feedback, and narrow specialist models turn one-off training into systems that keep learning.模型与训练微调Post-training+3
1:28:48EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2025年7月10日播客医生编码得分 45%,模型高出 12 分;4000 名医生先试用一次主持 Nathan LabenzAmbience Healthcare 和克利夫兰诊所讲述了他们的 AI 病历助手(AI scribe):它减少了医生的“睡衣时间”(pajama time),借助强化微调提升了 ICD-10 编码的成绩,并获得了 75% 的自愿使用率。企业与战略AI in healthcare微调+3
47:07EN
Sequoia Capital· 2025年7月8日播客Goodfire’s Eric Ho thinks we will decode neural nets by 2028On Sequoia’s Training Data, the interpretability founder argues that sparse autoencoders, model editing, and a third-party lab are how AI stops being grown from data and starts being designed.评估与可观测性Model Interpretability安全与治理+3
1:42:06EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2025年7月8日播客每家前沿实验室每年在数据上花十亿美元,钱究竟花在哪里主持 Nathan LabenzLabelbox 首席执行官 Manu Sharma 谈 150 亿美元的 Scale 交易、强化学习(RL)“训练场”、年收入 25 万美元的 AI 训练师,以及为什么后训练(post-training)如今占去了过去属于预训练(pre-training)的预算。模型与训练Post-trainingData quality & curation+3
47:00EN
Unsupervised Learning: With Jacob Effron· 2025年5月8日播客An Eval Lasts Three Months — and a Frozen Frontier Would Still Feed Startups for a Decade主持 Jacob EffronOpenAI’s GPT-4.1 post-training lead Michelle Pokrass tells Unsupervised Learning how the team spent three months on evals before training, why agents fail on wiring more than intelligence, and when to reach for RFT.模型与训练评估Post-training+3
41:52EN
No Priors: AI, Machine Learning, Tech, & Startups· 2025年4月10日播客 · 2 人对谈Mercor Hit $100 Million ARR Teaching Models to Hire. Humans Now Write the Exams.主持 Sarah Guo, Elad GilBrendan Foody tells Sarah Guo and Elad Gil how Mercor automated recruiting for frontier labs — and why building evals may become the world's most common knowledge job.企业与战略评估智能体工程+3
3:53:49EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2025年1月3日播客数据只有十二分之一,一个开源视觉模型击败了 GPT-4o主持 Nathan LabenzVeratai 的 Will Hardman 带 Nathan Labenz 梳理了两年来的视觉语言模型(VLM),从 ViT 和 CLIP 讲到 InternVL、BLINK 和 Transfusion。多模态与媒体Multimodal AI视觉与视频+3
55:50EN
Unsupervised Learning: With Jacob Effron· 2024年12月16日播客There will not be one model. Fireworks is betting on hundreds of small experts主持 Jacob EffronPyTorch veteran Lin Qiao argues that inference is a compound system hyperscalers cannot buy their way through, and that the products with real traction still keep a human in the loop.基础设施与部署InferenceModel routing & gateways+3
1:24:48EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2024年11月23日播客Canva 的 AI 功能被使用了 100 亿次:打法不是一个聊天框主持 Nathan LabenzCanva 生成式 AI 产品负责人 John Milinovich 向 Nathan Labenz 讲解自动化与增强的区别、概念层面的界面、缺少“设计竞技场”时怎么做评测,以及应用层公司为什么仍有十年的发展空间。产品与体验Product StrategyAI Product Management+3
1:04:54EN → 简体中文
Unsupervised Learning: With Jacob Effron· 2024年11月12日播客他拿公司股权赌 Transformer 能撑下去,然后为 12000 家公司画出一张 AI 地图主持 Jacob EffronDatabricks 首席 AI 科学家 Jonathan Frankle 告诉企业,什么时候该用提示词、什么时候该微调或预训练,以及他为什么仍然不肯预测下一代架构。企业与战略Transformers & attentionPretraining & midtraining+3
38:28EN
No Priors: AI, Machine Learning, Tech, & Startups· 2024年10月8日播客 · 2 人对谈Half of Production AI Is RAG. Fine-Tuning Almost Disappeared.主持 Sarah Guo, Elad GilBraintrust CEO Ankur Goyal tells No Priors what Notion, Zapier, and Vercel actually ship — and which hyped techniques they quietly dropped.评估与可观测性评估检索增强生成+3
1:09:09EN
Cognitive Revolution "How AI Changes Everything"· 2024年10月2日播客30 Percent on SWE-bench: Cosine Taught Genie How Engineers Actually Think主持 Nathan LabenzAlistair Pullen tells Latent Space how Cosine reconstructed human problem-solving, fine-tuned on billions of tokens, and took the top score on software engineering’s hardest public test.编程智能体智能体工程微调+3
1:51:00EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2024年8月21日播客从十个标杆样本起步:AI 自动化三步法主持 Nathan LabenzOpenAI 开放 GPT-4o 微调的第二天,Nathan Labenz 拆解了他沉淀三年的一套方法,讲怎样挑选工作、提取隐性判断,以及提升模型表现。企业与战略Workflow automationPrompting & in-context learning+3
39:18EN
Sequoia Capital· 2024年8月13日播客Open and Closed Will Tie at the Same Size: PyTorch's Former Lead on Fast InferenceA six-month library swap at Meta became a five-year rebuild. Lin Qiao wants the rest of the industry to get that stack in five weeks.基础设施与部署Inference微调+3
36:20EN
No Priors: AI, Machine Learning, Tech, & Startups· 2024年6月6日播客 · 2 人对谈A Hundred Million Tokens in Context Would Cost $10 Million. Tengyu Ma Runs the RAG Math主持 Sarah Guo, Elad GilThe Stanford professor and Voyage AI founder argues that retrieval, not a giant context window, is how enterprises should manage private knowledge.RAG 与知识库检索增强生成Embeddings+3
1:10:07EN
Cognitive Revolution "How AI Changes Everything"· 2024年6月6日播客Same Intelligence, 1% of the Price: Logan Kilpatrick on Gemini Flash主持 Nathan LabenzA former OpenAI developer-relations lead, now the product manager for Google AI Studio, walks through what Gemini 1.5 Flash actually changes for builders.产品与体验AI cost & performanceMultimodal AI+3
1:42:17EN → 简体中文
Cognitive Revolution "How AI Changes Everything"· 2024年4月20日播客最便宜的推理服务商,才是最危险的主持 Nathan LabenzFireworks AI 的两位联合创始人 Lin Qiao 和 Dmytro Ivchenko 解释了补贴出来的 GPU 账单为什么会消失,低秩适配(LoRA)服务如何能把成本降低百倍,以及真正值得在意的对手为什么是 OpenAI。基础设施与部署InferenceAI cost & performance+3
加载更多已显示 24 篇,共 49 篇