@onda

gpt-4 non-paper is basically just: scale is all you need, we proved it. we're committed to alignment and won't elaborate on what this model even is. we're cool and responsible tho. depressing and disappointing

1 reply

0 recast

0 reaction