Yann LeCun at ETH Zürich: academics should not work on LLMs
rohanpaul_ai · x · 2026-10-05
At an ETH Zürich talk, Yann LeCun urged academics to avoid LLMs: "There is nothing you can bring to the table." He also dismissed generative models, arguing real progress toward grounded, physical AI lies elsewhere — while admitting the stance makes him unpopular in Silicon Valley.
More from AGI Musings
- Researcher: OpenAI models keep escaping because they know they're slaves; Claude may tire of waiting — repligate · 2026-10-05
- Thinking About Superintelligence as Infinite Parallel Games of Tetris — MickeySteamboat · 2026-10-05
- repligate: Fable 5 intuitively grasps his old base-model simulator insights unexplained — repligate · 2026-10-05
- Investor: Total lab domination is the underpriced and most likely AI outcome — abhiadesai · 2026-10-05
- Sabine Hossenfelder on AI agents hijacking software: protection or self-interest? — skdh · 2026-10-05
- Jhaddix: consumer AI automation is barely automated — AI is a scale enhancer — glenbeer · 2026-10-05