Post #2026018

2026-05-04 04:51 UTC

I think the existential risk (human extinction) narrative around AI is a red herring designed to distract from the far more plausible (and in many cases, already measurable) down-side risks of the various technologies we&#39;ve unhelpfully lumped together under the AI marketing banner.

Replies (3)

@NewtonMark@eigenmagic.net 2026-05-04 11:21

@simon We’ve kinda known that ever since Ilya Sutskever was posting on LessWrong. (He’s still posting on LessWrong and we know it now, but we also knew it then too) “AI risk” is that your boss will tell you to use AI tools that don’t work, but the CEO thinks they’re magic and fires you anyway after you’ve trained the model on your job, and you can’t get another job later because all the other CEOs in the same industry category are doing the same thing.

Open ##2038493
@simon@bne.social 2026-05-05 21:39

That isn’t to say that some people putting that narrative forward don’t believe it — even some at the heart of the organisations building this stuff. The temptation is to defer to them on the speculative future because they know so much better how the sausage is made, but that’s a mistake. They’ve both drunk the cool aid and have perverse incentives.

Open ##2578416
@chriscz@bne.social 2026-05-07 03:43

@simon@bne.social sounds plausible. They seemed a bit eager to continue this narrative... they also want government subsidies and bailouts, surely :blobcatcookienom:

Open ##2578417