undefined | Better HN

0 pointsonlyrealcuzzo1mo ago0 comments

We've gone from yearly releases to quarterly releases.

If the pace of releases continues to accelerate - by mid 2027 or 2028 we're headed to weekly releases.

0 comments

But actual progress seems to be slower. These modes are releasing more often but aren’t big leaps.

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.

wahnfrieden1mo ago

GPT 5.3 (/Codex) was a huge leap over 5.2 for coding

rubicon331mo ago

Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5

minimaxir1mo 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.

j / k navigate · click thread line to collapse

0 pointsonlyrealcuzzo1mo ago0 comments

We've gone from yearly releases to quarterly releases.

If the pace of releases continues to accelerate - by mid 2027 or 2028 we're headed to weekly releases.

0 comments

rubicon331mo ago

But actual progress seems to be slower. These modes are releasing more often but aren’t big leaps.

gallerdude1mo 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.

wahnfrieden1mo ago

GPT 5.3 (/Codex) was a huge leap over 5.2 for coding

rubicon331mo ago

Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5

minimaxir1mo 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.

j / k navigate · click thread line to collapse