Yann LeCun(@ylecun)

@dylanbowmanSF Slide deck here: https://t.co/36v5ASQJZv

6.0内容质量
@dylanbowmanSF Slide deck here: https://t.co/36v5ASQJZv

TL;DR · AI 摘要

Yann LeCun分享了关于LLM局限性的观点及演讲资料，强调其无法单独实现人类级AI。

核心要点

  • LLM工具虽有用但非通向人类级AI的路径
  • 演讲资料包含完整技术论证（drive.google.com/lecun-20260911-eccv-keynote.pdf）
  • 视频版本可在YouTube观看（youtube.com/YannLeCunWorldModels）

结构提纲

按章节快速跳转。

  1. §核心观点

    Yann LeCun指出LLM无法单独实现人类级AI

  2. ·演讲资料

    提供包含完整论证的slide deck（drive.google.com）

  3. ›视频资源

    YouTube视频链接（youtube.com/YannLeCunWorldModels）

思维导图

用一张图看清主题之间的关系。

查看大纲文本（无障碍 / 无 JS 友好）
  • Yann LeCun技术观点
    • LLM局限性
      • 非人类级AI路径
      • 需结合其他架构
    • 资源链接
      • Slide Deck
      • YouTube视频

金句 / Highlights

值得收藏与分享的关键句。

#AI#LLM#YannLeCun#技术演讲
打开原文

Yann LeCun on X: "@dylanbowmanSF Slide deck here: https://t.co/36v5ASQJZv" / X

[](https://x.com/)

[](https://x.com/)

Yann LeCun on X: "@dylanbowmanSF Slide deck here: https://t.co/36v5ASQJZv"

  • ![Image 1: Avatar](https://x.com/ylecun) Yann LeCun @ylecun Sep 15 But there is a 1 hour talk, with results and evidence, that leads to these conclusions. Also, I am not saying that LLM-based AI tools are useless. They are very useful. We all use them. But, by themselves, they are not a path to human-level AI. Architectures that understand Show more 95 75 1K 63K
  • ![Image 2: Avatar](https://x.com/ylecun) Yann LeCun @ylecun Slide deck here: drive.google.com lecun-20260911-eccv-keynote.pdf 3:09 PM · Sep 17, 2026·23K Views 12 33 228 218
  • ![Image 3: Avatar](https://x.com/ylecun) Yann LeCun @ylecun Sep 17 Video of an earlier version of this talk Image 4 youtube.com Yann LeCun: World Models: Enabling the next AI revolution Talk given by Yann LeCun at ETH Zürich during "Frontiers of Embodie... 8 2 44 14K
  • ![Image 5: @nickshiva42](https://x.com/nickshiva42) Nick Shiva @nickshiva42 Sep 17 A question for Prof. LeCun: if we compare AI to a biological creature, does a biological creature learn to drive in 20 hours or in “1 billion years plus 18 years plus 20 hours”? Don't we create AI from a state of non-existence into a trained model in a few months? At the same Show more [](https://x.com/nickshiva42/status/2100700657800138881) 1 568
  • ![Image 6: @YTraonmilin](https://x.com/YTraonmilin) Yann Traonmilin@YTraonmilin Sep 18 "Any 10 year-old can clear up the dinner table and fill up the dishwasher without learning" Empty hypothesis implies everything. Interesting ideas nonetheless. [](https://x.com/YTraonmilin/status/2100835319813738738) 185

Log in or sign up for X

See what’s happening and join the conversation

Continue with phoneContinue with Apple

Continue with Google

通过 Google 继续操作 通过 Google 继续操作。在新标签页中打开

or

Log in with username or email