@cassidy@mastodon.blaede.family
Post #2935943
2026-05-29 01:03 UTC
RE: https://mastodon.social/@simple_sabotage/116655058624456803
I think this is part of what drives me crazy with LLM-powered code review: too often it feels like they’re following the US counterintelligence strategy for simple sabotage. Except I’m the target.
Replies (1)
-
@cassidy@mastodon.blaede.family 2026-05-29 01:12
To be clear, I think automated code quality tooling is great; CI with automated workflows, test suites, and linters are genuinely super helpful! But off the shelf LLMs reviewing my code might find a typo, but completely miss that I’m trying to solve the problem completely the wrong way. And so far, they have been *tedious* in a way (most) human reviewers aren’t.