AI usage guidelines across OpenWrt

[David Lang]

Josef Schlehofer wrote:

> Hi OpenWrt community,
>
> I think we should establish some common guidelines for AI use across
> the OpenWrt organisation.
>
> AI-assisted contributions are fine, but comments and code reviews
> should be written and verified by humans who understand and take
> responsibility for them.

I think that eliminating the use of AI for comments and code reviews is going to 
cost a lot of value. AI reviews should be able to be challenged (they may be 
wrong) but such reviews would go a long way towards easing the maintainer load 
by providing feedback on proposals automatically.

> AI-generated walls of text make discussions harder to follow and discourage 
> maintainers from participating.

they can, but they can also flesh out details that some people in the discussion 
may not be aware of (speaking as a human who has been accused of creating 'walls 
of text' myself)


David Lang