https://fisherdaddy.com/tags/richard-sutton/ Recent content in Richard Sutton on FisherAI https://fisherdaddy.com/images/papermod-cover.png https://fisherdaddy.com/images/papermod-cover.png Hugo -- 0.125.7 en Sun, 28 Sep 2025 15:48:41 +0800 https://fisherdaddy.com/posts/sutton-reinforcement-learning-vs-llm-bitter-lesson/ Sun, 28 Sep 2025 15:48:41 +0800 https://fisherdaddy.com/posts/sutton-reinforcement-learning-vs-llm-bitter-lesson/ 强化学习之父、图灵奖得主 Richard Sutton 深度剖析，为何当前火热的LLM（大语言模型）并非通往真正智能的正确道路。本文探讨了AI的本质、目标的重要性、对《惨痛的教训》的重新解读，以及通往通用人工智能（AGI）的“经验主义”范式。