Larry Ellison, co-founder and chief technology officer of Oracle, is a smart man. Larry Ellison knows databases. Larry Ellison knows how to sell databases. Larry Ellison built Oracle into an $450 billion company. Larry Ellison is an idiot.

Larry Ellison said he wants AI to be our policeman. Well, that’s not exactly what he said. Rather, he said AI is on the verge of policing our police. AI is also on the verge of policing all the rest of us, too. AI will be the always-on eye, and by dint of its unblinking-ness, none of us will dare do wrong.

To the extent that money talks, Ellison is fully on board with this seeming inevitability. Such technological omnipotence should be built, he believes, on Oracle’s technology.

That’s such a bad idea, even any current, credible AI doesn’t say that. And this should be end of matter.

If only it were that easy.

When Larry Ellison throws his wallet in with AI, what he is really saying is he wants AI to watch us all and judge our actions as in the right or in the wrong. Given the way AI “learns”, this means that he wants to rely on AI to look at us humans—how we treat each other—and synthesize an answer to what is right and what is wrong. Problem is, we don’t know right from wrong. Certainly not enough to program an AI that would judge us justly and with anything resembling equanimity.

As grim as such a technology sounds, we should thank Larry Ellison for raising this issue. It is a big and serious depiction of the human organism struggling to forge its conscience. It is, believe it or not, a sign of human progress.