UPD: The likeliest answer is that the problem is genuinely hard in the sense that it is hard to redistribute the wealth or resources towards the masses instead of AI companies.
I’m not too surprised. I should read him more! When I did experiments in Claude truesight of my comments/unpublished blog posts and memos, L Rudolph L came up surprisingly often. Perhaps the single most common result other than people who are locally famous like Scott and Kelsey Piper. Which is slightly surprising because I don’t think we read each other too much.
The likeliest answer is that the problem is genuinely hard
Yeah that makes sense. Probably my first guess too.
It seems to me that this dynamic was discussed in detail by L Rudolf L. However, if ASI alignment is wildly difficult, then we are either screwed or ban the ASI and use dumber AIs which end up being able to replace most human jobs. There also are many other posts, like By Strong Default, ASI Will End Liberal Democracy.
UPD: The likeliest answer is that the problem is genuinely hard in the sense that it is hard to redistribute the wealth or resources towards the masses instead of AI companies.
I’m not too surprised. I should read him more! When I did experiments in Claude truesight of my comments/unpublished blog posts and memos, L Rudolph L came up surprisingly often. Perhaps the single most common result other than people who are locally famous like Scott and Kelsey Piper. Which is slightly surprising because I don’t think we read each other too much.
Yeah that makes sense. Probably my first guess too.