2026-04-14 15:50 UTC
@ariaflame @hastar @Possiblydrew
Nobody claimed LLMs are sentient.
"just" is carrying a lot of weight there, the model weights encode why those next tokens are relevant and are the "magic".
Replies (1)
-
@ariaflame@masto.ai 2026-04-14 16:17
@hopeless @hastar @Possiblydrew So they magically create security risks, badly written code and duplicate functions all over the place.