Comment by rubicon33 12 hours ago But actual progress seems to be slower. These modes are releasing more often but aren’t big leaps. 4 comments rubicon33 Reply gallerdude 12 hours ago We used to get one annual release which was 2x as good, now we get quarterly releases which are 25% better. So annually, we’re now at 2.4x better. minimaxir 12 hours ago Due to the increasing difficulty of scaling up training, it appears the gains are instead being achieved through better model training which appears to be working well for everyone. wahnfrieden 12 hours ago GPT 5.3 (/Codex) was a huge leap over 5.2 for coding rubicon33 7 hours ago Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5
gallerdude 12 hours ago We used to get one annual release which was 2x as good, now we get quarterly releases which are 25% better. So annually, we’re now at 2.4x better.
minimaxir 12 hours ago Due to the increasing difficulty of scaling up training, it appears the gains are instead being achieved through better model training which appears to be working well for everyone.
wahnfrieden 12 hours ago GPT 5.3 (/Codex) was a huge leap over 5.2 for coding rubicon33 7 hours ago Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5
rubicon33 7 hours ago Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5
We used to get one annual release which was 2x as good, now we get quarterly releases which are 25% better. So annually, we’re now at 2.4x better.
Due to the increasing difficulty of scaling up training, it appears the gains are instead being achieved through better model training which appears to be working well for everyone.
GPT 5.3 (/Codex) was a huge leap over 5.2 for coding
Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5