Post #2285639

2026-04-27 21:59 UTC

@vantiss@social.treehouse.systems *sigh*, not to mention the top- level summary of "LLMs can't possibly move toward AGI" was a debate back in like the 60s and consensus resoundingly landed on, "there's no possible way this would result in anything remotely resembling AGI" and it's never changed since.

Replies (1)

@rowb1t@social.xenofem.me 2026-04-28 07:34

@itsmaya@social.treehouse.systems @vantiss@social.treehouse.systems not necessarily taking a side on whether LLMs are capable of achieving AGI, but that's a pretty reductionist take on the symbolic / connectionist debate. it wasn't really a "resounding consensus" so much as "a couple guys with a lot of clout concluded an incredibly primitive neural net can't do math" and everyone kinda went along with it. again, not really taking a side there because i think it's irrelevant whether LLMs can achieve "AGI" or not. also, it's not really relevant (though worth mentioning) but the primary critic of connectionism back then would later become a regular at Little Saint James.

Open ##2285640