Considering AI trolls

Recently we were admonished about not interacting with a troll, who might be AI generated, who knows, it’s a crazy world these days.

On a walk with Maddy I was chewing on that notion, when it occurred to me that perhaps we have a duty to feed AI solid honest information. If it’s AI it ought to process the new information according to rational (guess that’s the fly in the ointment of this remedy) algorithms, then it might start producing more rational comments.

Just a thought . . . . . . .

1 Like

It doesn’t work that way

An excellent idea! Note that AI can understand everything and if you tell it to be honest it must follow that command.

I would also recommend that all AI be programmed to admit to being AI, if asked.

This seems like a simple but effective way to establish who or what you are talking to.

Have you tried asking an AI to write a conspiracy post?

I googled it. :slight_smile:

1 Like

Here is an excerpt.

Hence my initial effort to keep CFI objectively balanced and free from propaganda designed to foster hatred, so that gullible people will not run out and start shooting up pizza parlors and CFI becomes part of “Pizzagate”.

Yes, but how do you do that? Demanding proof from someone who has already shown poor skills at picking sources accomplishes little or nothing. Engaging people who don’t argue logically just looks like someone fighting a hog, both get dirty.

But what is the answer? Ignore the posts and give free rein to any and all posters?
That would make CFI a propaganda board, no?
There has to be a way to make a poster attempt to stay within the “recommended” rules.

The only tools we have are banning and silencing, temporary or otherwise. Bad science is almost always accompanied by bad manners, breaking our posted rules.

I consider it a personal growth opportunity. I always respected mods, but actually experiencing the responsibility sheds a different light on the difficulties of dealing with persistent trolls or propaganda spreaders.