Post

Jake Wharton announced yesterday that all the Lysine projects (OkHttp, Okio, Retrofit, SQLDelight…) don’t allow contributors to use LLMs to generate code or write PR descriptions/comments.

Gunnar Schulze

@gunnar-schulze.bsky.social

16 Aug

Tbh, this move I heard of recently, to disallow LLM generated content in projects completely, is nonsensical. If the code is well-reviewed, line by line, and found to be good and valid, where's the difference? The problem is not AI per se but its usage.

Alec

@denmachi.dev

Opening a PR creates a burden on maintainers who need to do the labor to review it, as you say, line by line. And AI lets "developers" who have no idea what they're doing create giant walls of very plausible looking slop, creating a huge review burden.

16/8/26, 5:22 am0 reposts5 likes

Alec

@denmachi.dev

16 Aug

It's entirely cost benefit analysis. Are the benefits of the good PRs enabled by AI worth the cost of needing to sift through the slop? Or does it produce better results to just ban LLM generated code so maintainers can direct their finite time and energy elsewhere?