This credits the original publisher. Better Loop membership or a shared assessment is not implied.
The public work
Aider published code-editing benchmark results for an architect model that describes a solution and an editor model that converts it into file changes. The report compares paired configurations with solo baselines and discusses latency tradeoffs.
What to notice
Separate solving the problem from obeying a fragile edit format, then compare against the same baseline and acceptance checks.
Keep the context
These are historical, project-run benchmark results, not a universal model ranking or evidence of human improvement. The best reported configuration was also described as too slow for practical interactive use.
AI use: Documented in the source.
The report explicitly includes o1-preview as architect and o1-mini as editor in Aider, alongside other named pairings.
A useful public example is not an assessment of a reader, a publisher or a Better Loop member.
Authored practice suggestion
Try the idea. Check your own work.
Use material you are allowed to work with. This suggestion is preparation; it does not record a completed task or an improvement.
A check to adapt
Both attempts face identical checks; report any format or correctness gain together with the additional effort.
Ronacher published a tolerant XML parser after earlier parsing approaches failed. He describes directing the initial implementation, reviewing it, requesting further fixes and an extensive test suite, and releasing the resulting library.
Willison built a browser clipboard viewer while investigating rich-text paste. His public write-up links the code and conversation; the conversation shows requests to display available formats and then make long output easier to inspect.
GitLab’s postmortem reconstructs its January 2017 production-database outage, failed backup paths, and recovery from an earlier snapshot. It documents unrecoverable changes and traces follow-up work to monitoring, recovery testing, runbooks and operational ownership.