I have sometimes been accused of drinking the AI Kool-Aid.

And maybe that's right.

Certainly you can judge for yourself and feel free to tell me in the comments below.

But there's this distinction that my guest this week makes that I found it incredibly

useful, so I want to tell you about it as well. The idea is there are basically three schools

of thought around AI. First up, we've got the boosters, people who want AI to go as fast as

possible because think it'll be good, or for some intrinsic love of

speed. Then, on the other hand, you've got the critics. People who think that AI is running on

hot air, that it will shortly hit a wall for technical reasons, or because of the economics

not working out, or that sometimes even that talking about artificial

intelligence is to already have sort of fallen into a kind of marketing hype. But there's also a

third group, one that Garrison Lovely, my guest this week, calls the Warriors. And I am in that

group. I look at the graphs of how powerful these models are getting, and I don't

see us hitting a wall. I don't see the money running out, or even progress slowing down. So I'm not a

critic in this sense. But I'm not a booster either. I'm not happy about what I think automation will

almost certainly bring, which is the almost total disempowerment of the

global working class. And I've had this anxiety for years, but it's always remained slightly

nebulous until I read the book that we are discussing this week. Garrison Lovely is the author

of Obsolete, the AI industry's trillion dollar race to replace you and how to

stop it. The book goes systematically through the arguments of the boosters and the critics

alike. The book goes systematically through the arguments of the boosters and the critics

alike and argues that AI will be transformative, that that transformation will be largely quite bad,

...