9

Given the recent announcement Sites can now request to enable a banner to warn about their policy on AI-generated content and given our stance in How do we want to handle AI-generated answers?, it seems likely to me that we'd appreciate "a reduction in posts flagged for AI-generated content". (Happily, we haven't yet seen a large number of such posts).

I'll put two "voting" posts as Answers, one for each of the "yes" options:

  • **Reminder**: Answers generated by Artificial Intelligence tools are not allowed on [Site Name]. Learn more
  • **Reminder**: Answers generated by Artificial Intelligence tools must be cited on [Site Name]. Learn more

... but please add a separate Answer if you disagree with this idea for any reason!


Follow-up discussion on the "Learn more" Help Center page contents here: U&L Help Center article for our policy on AI-generated content

3
  • I like the idea of a title; thank you! I'm not very experienced in setting up Meta polls like this, so it's an adventure every time!
    – Jeff Schaller Mod
    Jan 5 at 21:11
  • The early/initial voting on these answers surprises me a bit, since the consensus on the previous genAI post seemed to allow for "answers OK with citation", but maybe the voters there were in favor of the "banned entirely" portion?
    – Jeff Schaller Mod
    Jan 5 at 21:13
  • If we're talking about banners, I think it should go without saying that any banner should reflect the site policy, and the question should be between adding a banner and not adding a banner. The two different phrasings shown here would represent different policies, and if we're discussing what the policy should be, then that should be clear from the post. Now the title asks "do we want to enable a banner", but the text effectively means this is a policy discussion. Not that the result appears to be changing, but in any case.
    – ilkkachu
    Jan 9 at 10:12

2 Answers 2

15

Answers generated by Artificial Intelligence tools are not allowed

I think we should have this banner added:

Reminder: Answers generated by Artificial Intelligence tools are not allowed on [Site Name]. Learn more

5
  • Pedantically, it's answers generated by Large Language Models. Practically, nobody knows the distinction.
    – user253751
    Jan 6 at 19:16
  • @user253751 Pedantically, the policy covers all AI generated answers. Although LLMs are the current thorny problem that the policy aims to combat. If other AI technologies become sufficiently advanced to generate answers then they remain banned until such time as the community feels specific technologies are acceptable. The primary issue with AI generated content is volume, and that will be the case for any AI for the foreseeable future. Jan 9 at 11:38
  • @PhilipCouling I think a large language model is by definition the only kind of model that can generate language. If someone somehow answers a question by the means of k-means clustering, it should be allowed.
    – user253751
    Jan 10 at 4:25
  • @user253751 my degree was in AI. No language generation AI has been around a long time, but it was rubbish by comparison. LLMs are a massive breakthrough, but by no means the definition of language generation. I'll stick with the idea that LLMs are the only kind of model capable of generating answers today. And no, the same problem would still exist for other models. The problem is one of easy volume of content generation. It's hard to see how any new AI would not bring the problem of volume unless that AI was bound to one specific piece of hardware. That's just the nature of software. Jan 10 at 10:02
  • 3
    This banner has now been enabled.
    – Sasha StaffMod
    Mar 14 at 18:43
-5

Answers generated by Artificial Intelligence tools must be cited

I think we should have this banner added:

Reminder: Answers generated by Artificial Intelligence tools must be cited on [Site Name]. Learn more

2

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .