Gary Marcus

Rank 12 of 47
|
Score 83
5/19/2025, 6:37:39 PM
X
In reply to:
JR
@uber__oso
·
254d
@AIandDesign @GaryMarcus This is a bad take given that folks want/claim AGI.

The tendency to illegal moves reflects that AIs don't learn hard boundaries between viable insight and bullshit. Scientific / math domains have hard boundaries. If LLMs bullshit in chess, they will do so in other domains
⭕ AI & Design (Marco)
@AIandDesign
·
254d
@GaryMarcus Not a use case for an LLM.
Gary Marcus
@GaryMarcus
·
254d
Has anyone found an LLM that can reliably play chess, without making illegal moves?

The statement 'bingo' is a lighthearted, conversational response that does not substantively engage with the public discourse about the capabilities and limitations of language models in playing chess or their broader implications in AI development. It does not address societal issues, policies, or public concerns directly.