miniblog.

I've been experimenting with LLM autoresearch. I prompted Sol to make difftastic faster without changing output on the test suite, and log everything it tried: https://github.com/Wilfred/difftastic/blob/c6e9c6fed4c71276a8b0e377c92695c3d929dc7f/PERF_RESEARCH_LOG.md It found some interesting performance tweaks, although it's too tolerant of complexity.

I've been impressed with code written by Fable in my testing: Difftastic: found small optimisations in a hot loop I'd already profiled extensively. Helped me prototype Dijkstra to A* too (hard to find a good heuristic). Garden: Found some real bugs in my simplistic typechecker.

One fun way of testing new AI models: take an existing codebase you have and just ask them to "review it and fix bugs". In principle this should find more issues over time as models get smarter. I've found a few bugs this way at least.