[daniel1212]

AI systems are learning to lie and deceive, scientists find - https://freerepublic.com/focus/f-news/4243717/posts

[RoosterRedux]

I read that the other day. Lying (for AI) will be much more difficult than just delivering the facts. And given the very competitive environment for AI platforms, a lying platform will get a very bad reputation and probably never recover.

Errors on the other hand are random and don't require a "motive."

I can see plenty of errors ("hallucinations"), but lying? That's pretty much on a par with Google's Gemini producing an image of a black George Washington. There's a leftie AI manager in the woodpile.

BTW, Gemini is pretty much dead because of that stunt.

[RoosterRedux]

I should have added, AI developers/programmers are idiots when it comes to marketing. Making and keepng their products marketable just isn't in their wheelhouse.

They are going to learn very quickly that they are now entering a world that demands a deep understanding of product marketing.

AI systems are expensive to develop, but everyone is doing it. That wasn't the world that Amazon and Microsoft grew up in.

If AI developers think they can develop platforms with a motive for spreading disinformation, they will quickly find themselves unemployable.