今天，我们不聊堆算力的“大力出奇迹”，而是要探索几条让AI变得更智慧、更可靠的巧妙路径。我们会看到，一个简单的“反刍”机制，如何让AI不再健忘；一场内部“辩论赛”，又如何教会它诚实；“专家分工”的智慧，怎样让它在变强的同时还更省钱。最后，我们还会探究AI是如何学会像高手一样“抬头看路”地做决策，甚至在数学领域领悟“功夫在题外”的道理。准备好了吗？让我们一起揭开这些最新论文背后的绝妙构思。00:00:37 AI的“反刍”，一个让它更聪明的简单魔法00:05:10 如何让AI变得更聪明，同时还不变坏？00:09:50 AI 进化新思路，从“大力出奇迹”到“聪明分工”00:14:55 高手决策的秘密，既要埋头拉车，又要抬头看路00:20:36 AI做数学，功夫在诗外本期介绍的几篇论文：[LG] Recirculation[Google DeepMind]https://arxiv.org/abs/2608.17981 --- [LG] Debate Training Reduces Reward Hacking in RLAIF[Google DeepMind]https://arxiv.org/abs/2608.17776 --- [CV] MoE-ViE: Mixture of Experts Vision Encoder for Efficient Image and Video Understanding[Meta]https://arxiv.org/abs/2608.17402 --- [LG] Q-Learning With World Models[Stanford University & Peking University]https://arxiv.org/abs/2608.17163 --- [AI] The Problem Is the Problem: Towards Scalable Mathematical Discovery[CMU]https://arxiv.org/abs/2608.16977 在小宇宙查看该单集文稿

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related