pull down to refresh
100 sats \ 0 replies \ @elvismercury 7 Dec 2023
My informal testing (having detailed conversations about complex topics) reveals that it's still a bunch less good than both GPT-4 and also Claude. This is a particular use case, YMMV, but it's the one I care about.
reply
21 sats \ 0 replies \ @Xlindo 7 Dec 2023
👀
reply