Do we want a banner warning about our policy on AI generated content?

Question

Since January 10th, sites have been able to request to opt in to have a banner shown on answers reminding answerers on their site's policy regarding AI generated content.

This leads to two questions:

• Do we want such a banner?

  • If so, which option? "Answers generated by artificial intelligence tools are not allowed on Code Golf" or "Answers generated by artificial intelligence tools must be cited on Code Golf"

• And, more implicitly, what is our policy on AI generated answers? Do we even have one? Do we need a site-specific policy?

  • Currently, from my understanding, moderators and users have been widely applying SE's general legal position that non-original works must be cited and credit given to the original author, and AI generated posts that do not meet this criteria are, by purpose of this more general policy, removed from the site. But what about AI posts in general, or those that do cite ChatGPT or similar?

Answer 1

I'll answer the questions in reverse order because the policy defines whether a banner is needed.

Policy

Currently there's no local code golf consensus on ai answers - a search for "ai" on code golf meta reveals no policy about banning usage of chatgpt and co. Further, there is no official network wide policy disallowing referenced ai posts:

With due consideration, we've decided no network-wide, general policy regarding banning ChatGPT, or other AI generated content, is necessary or helpful at this time.

Quote from an answer by Community Manager (CM) Slate regarding banning chatgpt network wide.

Noticed how I said that referenced ai posts are not disallowed (that is, ai generated posts that are properly cited as being made by ai are allowed by default). Posts suspected of being generated by AI without proper referencing are plagiarism and aren't allowed regardless of ai policy.

Therefore, the question becomes whether properly referenced ai generated posts are allowed. Personally, I'd lean towards allowing them because it's gosh dang impressive if you can get chatgpt to create a golfed answer that actually solves a challenge and abides by the competitive submission policy. Trust me I've tried from my experience it makes code longer.

Banner

Therefore, given the lack of policy, a banner isn't needed at this stage. How can people be warned about something that doesn't exist? If at some point there's a policy decision that ai generated posts that are properly referenced aren't allowed, then the banner would be a good idea.