Cas (Stephen Casper) on X: "Most of us have subtly different cruxy beliefs about AI that lead to some dramatic differences in opinion. As a result, I want to share 9 theses that I think explain almost all of my key thoughts on AI. [1.] The probability that some AI-triggered series of catastrophes will" / X

x.com · 426 words · saved by 1 readers

To view keyboard shortcuts, press question mark View keyboard shortcuts Home Explore Notifications Chat Grok Lists Bookmarks Communities Premium Profile More Post Neil Rathi @neil_rathi Post See new posts Conversation Cas (Stephen Casper) @StephenLCasper Most of us have subtly different cruxy beliefs about AI that lead to some dramatic differences in opinion. As a result, I want to share 9 theses that I think explain *almost all* of my key thoughts on AI. [1.] The probability that some AI-triggered series of catastrophes will result in the death of over 90% of the world's population before 2100 is probably ~5%. Corollary: Doom dominates AI risk calculus. (I don't love having this belief, but I have it nonetheless.) However... [2.] Other non-doom AI risks are extremely important too. Working to avoid them almost never conflicts and usually synergizes with work to avoid doom. [3.] The AI alignment problem (figuring out how to make an AI system consistently act in accordance with a d

Cas (Stephen Casper) @StephenLCasper Most of us have subtly different cruxy beliefs about AI that lead to some dramatic differences in opinion. As a result, I want to share 9 theses that I think explain *almost all* of my key thoughts on AI. [1.] The probability that some AI-triggered series of catastrophes will result in the death of over 90% of the world's population before 2100 is probably ~5%. Corollary: Doom dominates AI risk calculus. (I don't love having this belief, but I have it nonetheless.) However... [2.] Other non-doom AI risks are extremely important too. Working to avoid them al

related reading