AI disclosure fields

[avivace]

E.g. in #278 . We may want a way to precisely (self) declare if AI was used in the production of an entry and to which extent

[asiekierka]

In my opinion, separate AI disclosure tags for assets and code should be sufficient to cover real-world cases (there are, in general, people who don't mind AI, people who mind AI in the arts but not in code, and people who mind AI in both; there are some individuals with more granular distinctions but these are the three general sentiments I ran into).

But you could also leave it as a general descriptive field. Perhaps a more general (schema-bound, but extensible) list of "content warnings" to also cover works with other types of disturbing/sensitive material that people may wish to be warned about.

One issue that I've ran into when pondering about this is transitive dependencies: if an upstream project (some engine, music library, or compiler) is AI generated, should that lead to a disclosure of AI generated code being used? This question matters in particular as a vibe coded port of LLVM showed promise for a little while - would all ROMs made with it be considered disclosure-requiring by transitive property?

[avivace]

Thanks @asiekierka ! I've added some basic structure. Maybe not as simple as a simple tags attached to the entry but it should allow some 'precision' if wanted, or a simple yes/no.