HN saw Zig's anti-AI contribution ban as a maintainer-time policy, not a culture war

HN treated Zig's anti-AI contribution policy less as an ideological statement and more as a brutally practical answer to a maintainer-time problem. The submission took off because the project's rationale is unusually explicit: the scarce resource is not code, it is reviewer attention, and reviewer attention is supposed to turn newcomers into trusted contributors over time.

The policy itself is strict. No LLM-authored issues, pull requests, tracker comments, or even translation comments. Simon Willison's summary points to remarks from Zig Software Foundation VP of Community Loris Cro, who frames the review process as a long game. Zig does not mainly review PRs to squeeze in one more patch. It reviews them to teach people how the project works, build confidence, and slowly expand the pool of humans the core team can rely on. An LLM-generated or heavily LLM-shaped contribution can look polished, but it short-circuits that apprenticeship loop. The maintainer burns time reading it without necessarily gaining a contributor who can navigate the codebase independently next month.

That logic landed hard on HN. The top comments were not mostly moral panic about AI. They were closer to a resource-allocation discussion. If a patch is primarily machine-produced, one commenter asked, why should a maintainer spend time reviewing the submitter's artifact instead of generating a draft themselves and working from there? Another thread pushed back on the idea that Bun's upstreaming friction is only about the AI ban, arguing that some of the code has deeper language and architecture implications regardless of authorship.

The larger point HN kept returning to is that permissive AI policies and welcoming contributor policies may now pull in opposite directions. If a project accepts a lot of machine-assisted contributions, maintainers may become harsher with new contributors because review queues fill with output that does not reliably create future maintainers. Zig is choosing the opposite tradeoff: less throughput right now in exchange for a review culture that optimizes for human trust, context, and long-term project memory.

That is why the post resonated beyond Zig. HN saw a project spell out something many maintainers are already feeling: the argument is not only about code quality. It is about what a pull request is for.