>>5362
Rolf Degen should consider how he treats ants. Malice or even lack of sympathy are not necessary for human extermination to be convenient.

>>5362
>Machiavellians think others aren't "basically good and kind", have pathologically low time preference and tend to pursue future flexibility maximization
They sound more capable of empathizing with Clippy than other humans. Is this supposed to reflect badly on their ability to estimate AI risk?
Someone please post the PDF here. I can't get it through Sci-Hub.

>>5369
>>5364
The way you perceive AI risk varies with dark triad traits, people basically interpret AI risk to be omnipotent versions of themselves.

Nobody let big yud near a paperclip factory

>>5362
I hate this naive trend of attributing systemic issues to a lack of personal virtue. "It's the greedy bankers!" - change their fucking incentives, then!

Lmao

Well, I hate to say it but might is right. Humanity will either improve or be doomed.

>>5374

Exactly. Incentives matter a lot.

>>5374
I do hate that trend, but I don't really see it here.

>>5381
Framing the issue of AI safety as "of course the machiavellian STEM autists would project their own failings onto AI" frees critics from the burden of having to think about Omohundro drives and instead plays to their strengths of armchair psychologizing and sneering at bad greys.