A group of AI researchers from the Univ of Zurich just got caught running an unauthorized psych experiment with AI on a main communication platform

[LuvOrlando]

Well this is something, not exactly unexpected but completely unconscionable that a school thinks this is ok. I often wonder how often we all come in contact with this sort of thing, apparently there are no rules :/

Reminds me of the JK Rowling line from a Chamber of Secrets, "Never trust anything that can think for itself if you can't see where it keeps its brain."

https://www.nbcnews.com/tech/tech-news/reddiit-researchers-ai-bots-rcna203597

How long has this been around? Only a year or so... yikes

 

[Mommb]

I understand why this was wrong and why the moderators on the chat board were upset about it, because it was against the terms of the site. However, it does not appear that the bots actually hurt anyone. The experiment was in a forum dedicated to explaining positions, and that’s what the bots did.

It is common for online posters in anonymous forums to say things they wouldn’t say when talking to someone they know, to stir up debate or just to see what happens. How is using a chat bot in this case significantly different from a human poster saying something to see what happens?

I am more concerned when AI bots are used to provide specific information, as now happens with some prompts I put into Google. If the information the AI gathers (for example, about health care or legal issues) is incorrect, it can lead to real harm.