logoalt Hacker News

0c3ca83 • today at 5:05 AM • 1 reply • view on HN

They seem to believe that one. I suspect extreme harm to human well-being is far more likely.


Replies

antonvs • today at 7:09 AM

I'm skeptical that they really believe it - at least, not the people at the top.

There's no question that AI can, and almost certainly will, be used to do great harm, but the idea that AIs are just going to up and decide to kill us on their own is unlikely fanfic at best.

And at worst, it's projection by people who are really telling on themselves about what they want to do to anyone they consider either useless or a threat. They're also thinking of the consequences of the way they plan to use AI themselves.

They read or watched "Ender's Game" and thought it was a model for how to deal with competition. Or, as we see with Thiel for example, they read Lord of the Rings and thought "Sauron did nothing wrong." Or in Musk's case, substitute Sauron with the guy whose salute he likes to copy.

I realize neither of those people are really central to the AI race, but they're symptomatic of the thinking of the cult. In the 1980s, the people who deluded themselves that they were "masters of the universe" were on Wall Street. Now, they're in Silicon Valley.

In fact, if AI were to turn on its creators, we might end up with the best of all possible worlds.