Raku / Rakudo adapt and adopt the LLVM AI Tool Policy

[librasteve]

Some time back, it was noted that new Raku ecosystem contributions were over represented by AI intensive vibe coded slop from one or two authors.

This proposal is twofold.

1. The Raku Steering Council reviews and adopts a policy identical to the recently proposed LLVM AI Tool Policy for all contributions to raku and rakudo and all bundled modules.

2. The fez upload process (and other contributions tools) require a module authored to self declare that their code adheres to the policy. Ideally this would be via a new tag in META6.json. Violators could then have modules tagged in eg raku.land as a warning / sanction.

Feedback and thought welcome.

[tbrowder]

First, the acronyms LLVM and AI should be spelled. Next, the lead in should say something like whose policy we are talking about. I suggest something like:

The Raku Steering Council hereby promulgates its policy on the
use of Low Level Virtual Machine Artificial Intelligence, now
commonly known as LLVM AI, for generating code or text for 
submission to the Raku organization.

@codesections, being a lawyer, knows how to word that better.

Also I would like to see a link to the LLVM project on GitHub.

It would be good to specifically mention the several models known to be used already (with links): ChatGPT, GitHub CoPilot, etc.

[lizmat]

See also: #491

[tbrowder]

I'm a dummy, when I started my response I didn't realize that link was the actual LLVM organization's policy, so take my input with a large dose of salt

[jonathanstowe]

Broadly in favour, though the two points may want to be decoupled to some extent as the first is quite clear and speaks to the time and effort of the community as in the policy, whereas the second is a bit more nuanced in my mind as the upfront "cost" to the community isn't as clear.

[librasteve]

OK - seems that there is a consensus to go for step 1 in relation to PR submissions to raku repository code. I will seek to add step 1 to the next RSC agenda...