By Katja Grace, April 1 2021
Today we are sharing with our blog readers a collection of yet-to-be-published drafts, in the hope of receiving feedback. We are especially looking for methodological critique, but all comments welcome!
Human-level performance estimate (Katja Grace)
How much hardware will we need to create AGI? (Asya Bergal)
Historic trends in AI Impacts productivity (Daniel Kokotajlo)
Analysis of superbombs as a global threat (Katja Grace)
Looking at your list of objects, used to estimate how much hardware will be needed to create AGI, I am worried that there may be a problem, related to something I've heard of but didn't quite understand, called "reference class tennis". To address the problem, I suggest adding a tennis racket, a tennis ball, and a tennis court to the list of objects. This should also help to make the result more statistically significant, by increasing the sample size.
Announcement: "How much hardware will we need to create AGI?" was actually inspired by a conversation I had with Ronny Fernandez and forgot about, credit goes to him for the original idea of using 'the weights of random objects' as a reference class.https://i.imgflip.com/1xvnfi.jpg