lonjil.net
very sleepy
👤 Joined on 11 May 2023
Followers
Following
Total Posts
@lonjil.bsky.social@lonjil.bsky.social → @lonjil.xyz@lonjil.xyz → @lonjil.netNot Found
Not Found
Not Found
just--bruno.bsky.social
aliciasamm35.bsky.social
printplayprotect.bsky.social
tovofrttb149.bsky.social
spoont51.bsky.social
Not Found
Not Found
lawoffairness.bsky.social
rifka.bsky.social
Not Found
Not Found
dhjouxfp1800.bsky.social
thefinancenewsletter.com
charlesreynaart.bsky.social
us-political-news.bsky.social
Not Found
Not Found
followblue.app
imagesbyjonevan.bsky.social
Not Found
sylvaners.bsky.social
roddsdad.bsky.social
If you can only make decisions by putting the probabilities into a spreadsheet then you have no choice but to pull numbers out of your ass or you’ll never be able to decide what the most important cause in the world is
Show PostI mean, why not 9%? Why not 20%? ’More than 10’ is just so clearly pulled out of their ass
Show PostThat is to say, the motivation to express a p(doom) comes from the idiosyncratic epistemic / ethical decision-making framework, rather than the availability of sufficient evidence to estimate such a number, whatever it may mean.
Show PostWhat makes it click is to understand that Rationalist/EA thought says you should allocate resources according to expected values, and so you need probabilities to plug into the expected value formula. To decide what to allocate to AI Safety, you need a number for the spreadsheet. That’s p(doom).
Show PostIt really doesn’t make sense without understanding that the speaker has a strange worldview that requires numerical probabilities assigned to everything so that he can perform the necessary calculations to determine what’s important and what’s not. Most people don’t have this worldview, though.
Show Post