Haiku should rethink that AI rule. Put very high code quality gates in place, sure, but Haiku in particular is a very low-velocity project and AI could help them so much. There is no reason you can't produce high-quality code with AI conformant to your style guidelines and structural preferences, it just takes a lot more effort than the slop.

▲

LeFantome 3 hours ago | parent | next [-]

I imagine that their issue is reviewer bandwidth.

If all submissions were high-quality, it would probably work. But that is probably unlikely.

Maybe the project could host a searchable directory of AI contributions hosted at external GitHub repos or something.

	▲	SyneRyder an hour ago \| parent \| next [-]
		My understanding is that the AI policy is more to do with issues of copyright and code provenance. They want to make sure that all of their code is cleared with no risk of accidentally using 'stolen' code or GPL code repeated by an AI. I can at least understand the policy. I think there's other opinions around AI use there as well, but I'll back off from that. I didn't mean for my post to become top voted, when we should be celebrating Haiku reaching Beta 6. I'm just disappointed that for me, R1/b6 has been a very big regression, right at a time when I've been using AI to make all the software I write cross-compatible with Haiku via Go & SDL, and even use AI to write drivers so more of my hardware works on Haiku... and instead, now I can barely even boot the system.
	▲	hvb2 3 hours ago \| parent \| prev [-]
		> I imagine that their issue is reviewer bandwidth. This feels to me as the real bottleneck in general. Maybe along with thinking through what to code/prompt. I don't need more tokens or automated generation of PRs, I'm already at capacity for the care and understanding I want to have for code I own.

▲

cardanome 2 hours ago | parent | prev | next [-]

No one is trying to get rich from working on Haiku. They are doing it for fun, as a passion project. Sure AI would help with velocity but velocity is not the only goal. Journey before destination.

Also: Open source is about relationships.

Reviewing code sucks. However at least reviewing code from humans allows you connect and learn from each other. You get to mentor someone. Build a relationship and trust. They learn and grow and maybe someday can also review other people's code.

Nowadays with lots of low effort merge requests and spam it is already easy to burn out. Now add AI. Yeah, no. Why should anyone review AI slop for free. Why? It is the prefect solution to burn your contributors out.

▲

shevy-java 3 hours ago | parent | prev [-]

I have no qualitative opinion as to whether the Haiku team should reconsider the AI rule or not, but you select about just one possible advantage. There may be disadvantages using AI. One that keeps on happening is that real humans use AI slop spam to pester other humans with it.

> There is no reason you can't produce high-quality code with AI conformant to your style guidelines and structural preferences, it just takes a lot more effort than the slop.

It's not just code though. There are humans who deliberately and actively use AI slop to spam people. I saw this first on gobolinux, then on prawn, then on the ffmpeg mailing list. There are really good reasons why many humans dislike the proliferation of AI slop spam.

> Haiku in particular is a very low-velocity project and AI could help them so much

I am not entirely certain about that. If AI were to work so well, Haiku-AI could have already been insta-spawned and auto-fixed everything. Clearly this does not work, so I am unconvinced that AI is the final solution to everything - at the least right now.

What is also unfortunate about Haiku is that things such as ruby do not work by default. Ruby even works on windows, though on Linux it is much more convenient to use.