Should AI systems have to identify themselves?
A short post to hear what people think of the policy proposal to have any interaction with an AI (eg chatbot) include an overt notification that one is indeed interacting with an AI.
Clearly, one wouldn’t want a frequent pop up in video games but in the world of providing services to consumers, such a thing could help reduce confusion among clients/users.
Alternatively, perhaps it isn’t that helpful or might be treated like the cookie notices we get and just click accept.
What do you think?
One problem is that there’s no that there are a lot of shades of grey in between. If I used an AI grammar checker to answer your question, do I have to mark my answer as AI generated?
If I let ChatGPT rewrite my answer to sound smoother, do I have to mark my answer as AI generated?
If I let ChatGPT generate four answers and chose the one I like best to post, do I have to mark my answer as AI generated?
(1) No. (2) “Contains AI-assisted content.” (3) “Contains AI-generated content.”