News for hackers. Stuff that matters.

A comment on Improving 15 LLMs at Coding in One Afternoon. Only the Harness Changed. Parent comment

Re:Improving 15 LLMs at Coding in One Afternoon. Only the Ha... by mycall on Thursday February 12, 2026 @11:17AM (#46990649)
If I remember, both Claude Code and OpenAI Codex "harnesses" improved themselves now.

OpenAI used early versions of GPT-5.3-Codex to: debug its own training process, manage its deployment and scaling and diagnose test results and evaluation data.

Claude Code have shipped 22 PRs in a single day and 27 the day before, with 100% of the code in each PR generated entirely by Claude Code.

[ Parent | on HN ]
Re:Improving 15 LLMs at Coding in One Afternoon. Only the Ha... by withinboredom on Friday February 13, 2026 @05:23PM (#47008621)
> with 100% of the code in each PR generated entirely by Claude Code.

You can tell...

[ Parent | on HN ]