Dan Romero
@dwr.eth
@pushix wdyt about o3
1 reply
0 recast
10 reactions
Nicholas Charriere
@pushix
early but super excited tbh Seems really good at reasoning and coding. Need a bit more empirical data (will make my own benchmarks) and the next week should confirm / deny hype. I don't understand why they launched everything else though. If this is truly as good as they claim it should be the one and only thing they're talking about all the time. 72% on SWE bench is serious stuff. Going to try it in Srcbook :D
1 reply
0 recast
6 reactions
Jeff Feiwell
@hyper
Better than 3.5 ?
1 reply
0 recast
0 reaction
Nicholas Charriere
@pushix
Idk yet I need them to give api access so I can plug it into srcbook
0 reply
0 recast
2 reactions
