@GaelVaroquaux There have definitely been wild claims on either side. But in some sense, we already have "AGI": I suspect LLMs are better than most people at most problems that can be solved on paper. At the same time, I do think a paradigm shift is needed for certain key abilities.
@andrewgwils Well, on the other hand Sam Atlman and the likes said that we would have AGI by 2025 and that it would be dangerous for humanity
