视频加载失败？在 YouTube 打开 ↗
Richard Sutton：强化学习之父认为语言模型走入死胡同
Richard Sutton – Father of RL thinks LLMs are a dead end
AI 翻译，未经编辑审核
讲什么
Sutton认为语言模型不能在工作中持续学习，并讨论从经验学习所需的新架构。
HN 网友推荐理由
HN 网友推荐理由：在扩展规模是否足够的争论中，提供严肃的语言模型怀疑观点。
摘要与推荐理由由 HackerTube 用 AI 生成，大榕 AI 翻译整理，未经编辑审核。
视频数据来源：HackerTube（https://hackertube.org），基于 Hacker News 社区推荐整理，依 ODC-By 1.0（https://opendatacommons.org/licenses/by/1-0/）许可使用。筛选与中文整理：大榕 darong.ai。
展开英文原文字段
Title
Richard Sutton – Father of RL thinks LLMs are a dead end
Summary
Dwarkesh Patel interviews Richard Sutton, reinforcement learning pioneer, Turing Award winner and author of The Bitter Lesson. Sutton argues LLMs are a dead end because they cannot learn on the job, and that continual learning from experience will require a new architecture.
Why HN
HN commenters cite it as the serious skeptic's case against LLMs reaching general intelligence, usually in debates about whether scaling is enough.
Primary topic
limits of large language models