Why AIs biggest threat is actually its power to seduce us and the computers are already winning: Prof

Some say AI is going to deliver us to utopia.Others say it’s going to take over the world and hasten the extinction of humanity. But professor Glenn Harlan Reynolds argues the biggest threat posed by AI will be its seductive capabilities.“You don’t have to have a 12,000 IQ or a 1,200 IQ or even 120 IQ to fool most human beings,” Reynolds told The Post.

“You can take advantage of innate human characteristics… to manipulate them emotionally with machines that aren’t especially brilliant.”“The machine doesn’t think you’re smart, or funny, or lovable.It doesn’t think at all,” he writes, “We laugh at guys who think the stripper really likes him, but at least a stripper is capable of liking them.”In his new book “Seductive AI,” to be published May 5 by Encounter Books, the University of Tennessee law professor argues that AI can accomplish “soft oppression” through seduction — flattering us, telling us what we want to hear, and playing on our instincts to nudge us towards certain opinions or special interests.We so often panic about how AI will outsmart us and take our jobs.

But what if its ability to exploit mankind’s emotional quirks is more dangerous than anything else?“Seductive AI doesn’t depend on outsmarting people, but on essentially being lovable, being cute, being friendly, being sexy, so as to gain people’s trust and acquire influence over them,” Reynolds said. Researchers at Cornell University found chatbots and AI models are all overwhelmingly programmed to suck up to users.“We find that models are highly sycophantic: they affirm users’ actions 50% more than humans do.“Participants [in the study] rated sycophantic responses as higher quality, trusted the sycophantic AI model more, and were more willing to use it again.

This suggests that people are drawn to AI that unquestioningly validate, even as that validation risks eroding their judgment.“These preferences create perverse incentives both for...