Yann LeCun at ETH Zürich: Don't Work on LLMs or Generative Models If You Want Real AI Progress
CurieuxExplorer · x · 2026-10-07
Speaking at ETH Zürich, Yann LeCun urged academics to stay away from LLMs: "There is nothing you can bring to the table." He argued that anyone interested in grounded, physical AI should avoid both LLMs and generative models — "this does not make me very popular in Silicon Valley."
He also noted that a four-year-old takes in as much visual data as the text used to train the biggest LLMs, yet already understands basic physics that current AI does not — consistent with his long-standing bet on world models over the text-centric paradigm.
More from AGI Musings
- Writing a precise spec IS half the work — AI helps those who already specify well — sujingshen · 2026-10-07
- LLM agents pick worse items from trusted sites; same price cuts Booking.com bias by 28 points — rohanpaul_ai · 2026-10-07
- Turing laureate Patterson: AI will replace all jobs by end of 2030, knowledge work to drop below 1% of today's price — davidpattersonx · 2026-10-07
- Blogger Slams Anthropic's Effective Altruism Culture: Ideology Isn't Humanity's Consent — VraserX · 2026-10-07
- Grady Booch: a non-small number of mathematicians are deep in soul-searching — Grady_Booch · 2026-10-07
- eigenrobot: ethical norms work like evolved organisms, not rational machines — eigenrobot · 2026-10-07