she "disagrees with catastrophic framings of AI risk."
that's not very consistent with my understanding of the words "endorsed IABIED" from OP
Feels true to me, but what's the distinction between theoretical and non-theoretical arguments?
Consider the mythological case of the calculation if the atomic bomb would ignite the atmosphere. I guess the concern guided the "policy" to perform the calculation. And if it came out as 50% of omnicide, atomic bombs would be prevented, despite the lack of a spectacular warning shot.
Policy has also ever been guided by arguments with little related maths, for example, the MAKING FEDERAL ARCHITECTURE BEAUTIFUL AGAIN executive order.
Maybe the problem with AI existential risk arguments is that they're not very convincing.
Common rationalist take is that people used to really believe their religions (and now it's fake).
Somehow I can't not doubt that 1st century people unironically believed the Bethlehem census story. They would be familiar with state logistics of the time!
They must've been like yeah we made it up for Messiah lore lmao
This seems pretty useless.
Step 1 won't work for distinguishing good from bad, current tech is not capable of this.. (I guessed that it would be biased to be negative since it's kinda suggested in the prompt, but from other comments it seems it still glazes)
As pointed out by other comments, step 2 would reject much/most real knowledge progress.
"Support" is kind of weak. To make it like the CAIS statement, maybe "I largely agree with IABIED". Or "I ~agree with IABIED". Or "I agree* with IABIED [bring your own footnotes]" 🙂
"Statement of ~agreement with IABIED"
I never got that cause is deciding to smoke much of an update after you already detected an urge to smoke? edt looks simpler so it should be correct
real, thanks. both look pretty garbage I guess no choice but to drink it
I used that once and it didn't work, aligned-by-default universe
that's a trick to make me be like them!
(I listened to some of that michael huemer talk and it seemed pretty dumb)
I started using Forfeit from this post, and it worked amazingly for me, but now it turned into subscription AI slop. Is there an alternative?