Nathan Barnard
nathan-barnard.bsky.social
Nathan Barnard
@nathan-barnard.bsky.social
Helping people is good |
Empirical evidence is important |
Blog at http://thegoodblog.substack.com

It's so good! (and on audible if that's signfgiant for you)
December 10, 2024 at 7:59 PM
have you read Deng Xioping and the transformation of china
December 10, 2024 at 7:30 PM
Yeah i think some of the darker findings from anthropolgy are so interesting.
December 10, 2024 at 5:49 PM
i would love a hot take
December 9, 2024 at 6:35 AM
I'm reading "By Steppe, Desert and Ocean" at the moment and "Blood and Iron" - both excellent (and the first is free on audible although that may be less relevent you than it is to me
December 8, 2024 at 7:01 PM
(and good practice particuarly when one has a strong emotational reaction)
December 2, 2024 at 12:54 AM
I don't think that this is equvilent - to be very clear - but I think in hindsight the social democracts were wrong not to play dirty in weimar.

Again, I think that this action is, in expectaton, one of the worst a modern president will take, and i'm steelmanning because it is good practice.
December 2, 2024 at 12:54 AM
My gut reaction is that I strongly agree and think that this is horrific.

I slightly wonder if there are good tit for tat game theroy reasons to do this.

I think almost certainly not, but that's my steelman of the case for it.
December 2, 2024 at 12:52 AM
FWIW I think your shutdown problem work is pretty ambitious (although perhaps should be viewed as a special case because it's applying results in mature field to a large problem in an immatuture one, which perhaps gives us a priori reason to think that one can gain traction on the largest problems)
December 1, 2024 at 10:45 PM
My guess is that this is also feild depdendent, and the younger the peak of the feild is the more ambitious one should be early on. Of course there's also a training compoant and you have more inside knowledge on this than me.
December 1, 2024 at 10:45 PM
I think in areas where it's difficult, aiming for results which are unlikely on priors is socially harmful. Social psychology is the obvious area where this is the case.
December 1, 2024 at 10:45 PM
I think a subtility here is around how easy false positives are to detect. In formal subjects, where I think it's roughly easist, I think it's worthwhile being very ambitious because it's easy to verfiy.
December 1, 2024 at 10:45 PM
If you could know ex ante of course how likely you were to do paradigm shifting work then that would be great, but that's not obvious to me (perhaps outside of pure maths but even then most fields medalists were not IMO medalists.)
December 1, 2024 at 10:16 PM
But like the EV seems worth it? Academia is extremely heavy tailed, and although plausilby for induviudal students this is bad advice with respect to their prudential interests, it could still be socially valueable for most people to take these longshots.
December 1, 2024 at 10:16 PM
somewhat forshadowing a more serious post on christianity
November 29, 2024 at 12:05 PM