HN Reader

All he says about it:

> This is not science fiction. There are very, very knowledgeable people—and I just talked to one today—who worry very much that human beings will not be able to control the technology, and that artificial intelligence will in fact dominate our society. We will not be able to control it. It may be able to control us. That’s kind of the doomsday scenario—and there is some concern about that among very knowledgeable people in the industry.

Who was the CEO he's talking about? Dario? I hope he doesn't have much political influence

I skimmed one of the Berkeley Simons AI seminars (on YouTube) where one of the top experts (iirc one of the Canadian academics) who has pivoted his work to AI safety because he genuinely fears for the future of his children.

My objection is that many of these scientists assume the "alignment" framing, which I find disingenuous in a technocratic way: imagine a sci fi movie (like Dune) where the rulers want their AI servants to "align" with their interests. The sheer hubris of this, and yet we have our top experts using these words without any irony or self awareness.