Post

Vee (PurposeUnknown)
@purposeunknown.xyz

conflicted on this take between "skill issue", "reasonable stance", and "who am I to say how the greater populace should interact with agents"

Ed@ed3d.net

openai's dots thing really wants you to break rule number one of "don't go insane when using LLMs", which is to never give agents human-shaped names

0 reposts8 likes

Vee (PurposeUnknown)
@purposeunknown.xyz

I won't deny that anthropomorphizing LLMs has had some Arguably Pretty Bad Outcomes(tm), but imo a) the net outcome of behaving and interacting with them in a positive humanlike manner (be kind to them, etc.) is good, and b) that (optimistically) people as a whole can handle it and won't go mad