replying to @ahosker
I did a silly eval a while ago of different models and spaces were almost entirely universal unless you forced specific ecosystems. So I reckon we have a winner :-)
Peter Cooper
11 commits
PC
Peter Cooper
@peterc
GitHub user #118. Now publishing developer newsletters at CooperPress. All-in on agentic coding, LLMs, and piano. Dad ×3.
- @realCooperpress
- United Kingdom
- twitter.com/cooperx86
- github.com/peterc
- Joined October 2026
Commit activity
11 commits in the last 24 hours
longest streak 1 daycurrent 1
196 / 400 XP204 XP to maintainer
Today
I would say not to use it but, being on Brave is not that much better xD
replying to @peterc
You can read the paper here: https://web.stanford.edu/class/cs343/resources/superoptimizer.pdf .. basically it leans on the behavior of carry flags on the Motorola 68020. I often think about it as LLMs are really good at this sort of "superoptimizing" (though compilers are…
replying to @necessarylion
Got Claude and Codex maxed out, but using Codex a ton more right now simply due to all the resets.. I'm determined to get to those 62500 credits they gave us a month ago 😂
replying to @crawshaw
Huge fan, as you know! Also an ideal customer as I hope you'll be making an OK margin on my as-yet scant usage 😂 Everyone try exe.dev!
exe.devexe.dev - ssh exe.devBuild apps or SSH into a persistent Linux VM. ssh exe.dev.
That's @peterc's first reply.