Somehow every time I see animals solving a problem, it reminds me of @ylecun.bsky.social perspective. Two questions: Can a reasoning model (e.g. O3) solve it within a sandbox (I think it is harder than ARC-AGI)? And can multiple agents or robots collaborate to tackle the problem together?”
