Why does chess require reasoning? Do all of the these [1] "reason"? ChatGPT-4 is supposedly rated worse than 500, in this list (1400 or so, although I think a recent update improved it a bit).
Within the domain of chess, searching the domain of possible future positions is synonymous with reasoning. It requires an explicit understanding of latent board representations and an explicit understanding of possible actions and the consequences of those actions.
Whether ChatGPT has any of those things is questionable. From what I’ve seen, it has an unreliable latent representation, an unreliable understanding of possible moves, and a positional understanding that’s little better than a coin flip.
> Within the domain of chess, searching the domain of possible future positions is synonymous with reasoning.
So, with this definition, these chess engines already exhibit (fairly substantial) reasoning? Or, are you saying it would be required, in the context of an LLM?
[1] https://ccrl.chessdom.com/ccrl/4040/