Nick_Tarleton comments on On the Power of Intelligence and Rationality - Less Wrong

13 Post author: alyssavance 23 December 2009 10:49AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (187)

You are viewing a single comment's thread. Show more comments above.

Comment author: Nick_Tarleton 23 December 2009 08:08:24PM *  2 points [-]

A superintelligence trying to figure out how to take over a country would probably create a public enemy and invoke as much Nationalism as possible, and publish lots of propaganda, and indoctrinate the youth, and establish a cult of the leader, and so forth.

If true, this seems to support the post, since they got the right answer without being rationalists.

Acting rationally means using the means that will accomplish your desired ends most effectively.

Yes, but (not being unitary agents and all that) even the most sincere rationalist-wannabe human probably won't do so, or even use the most effective humanly-achievable means; and one merely trying probably won't use the most effective means they can. (I know this is blatantly obvious, but I suspect repeating it is valuable.) Also, the valley of bad rationality may extend a damned long way in some domains.

If the Nazis had had their crazy values but actually been rational and carried out their war differently (less aggression towards Russia, not dragging the US in, delaying the Holocaust until they could actually afford the military resources for it), they could have been a whole lot more successful than they actually were. So let's be glad they underrated rationality.

Agreed.

Comment author: Psychohistorian 23 December 2009 08:42:49PM *  4 points [-]

I hadn't recognized the survivorship bias issue here. There are a lot of crazy groups that try to take over countries. Every now and then, one succeeds. This doesn't imply that being crazy will consistently lead to effective coups.

ETA: Arguing, "The Nazis succeeded despite inadequate rationality," is of limited value as evidence. There are a LOT of political groups out there, virtually all of them are fairly irrational in some respect, and a few are bound to succeed. Thus, arguing, "Rationality isn't important to success, because so-and-so succeeded," is only useful if they have been consistently beating out more rational groups.

On an individual level (and I'll put this in more detail in a separate thread) Tom is absolutely right that there are more important factors in individual success than mere rationality. This just fails to come across in his post, at least on my reading of it. The Nazis are, I think, a substantial distraction if this is indeed the point.

Comment author: brazil84 27 December 2009 02:20:45AM 0 points [-]

I agree. Besides, the rational thing to do might be to steer clear of putsches, coups, and the like. And even if you are the democratically elected head of a state, your chances of being assassinated are pretty good.

Comment author: alyssavance 24 December 2009 02:36:45AM 0 points [-]

See http://lesswrong.com/lw/1ko/on_the_power_of_intelligence_and_rationality/1d5t . If rationality is an important factor in success, you'd expect pretty much all of the successful groups to have above average rationality.

Comment author: orthonormal 26 December 2009 07:52:21PM *  0 points [-]

I'd expect them to be above-average within the reference class of "cabals attempting to overthrow the established political order", not the reference class of "people in general" or "organizations in general". And I'd expect the average rationality level of that first category to be pretty low.