mfs be like "AIs don't have internality, they're just imitating human introspection from the training data" and here ChatGPT starts explaining the meaning of a token to me as though I'm also an AI who has token qualia and can't visually perceive whether the token is a word or not
retweets 0
likes 0
substance 4 · opinion · rank 60 · ChatGPT exhibits its own distinct internality rather than simply imitating human introspection.
view on X ↗