Nick_Tarleton comments on Beware of Other-Optimizing - Less Wrong

79 Post author: Eliezer_Yudkowsky 10 April 2009 01:58AM

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

Comments (117)

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

Comment author: byrnema 10 April 2009 12:26:50PM *  8 points [-]

"Different things work for different people."

Living life, getting through the day, is obviously an enormously complex process. Whether we are rational or irrational, we make decisions based on a large number of short-cuts. These "short-cuts" have evolved over time and have their origins in our routines, our values, etc. However, since they don't completely capture our full decision-making system (i.e., don't reproduce every time the decision we'd make if we had enormous time and energy to decide each one), they introduce certain inefficiencies (or even inconsistencies).

I think that a lot of advice represents "hacks" for getting around certain built-in inefficiencies. It would be enormously trouble-some (and arguably counter-productive) to rewrite your whole meta-code, but a patch is worth adding, if it works.

Thus I think the reason advice works for some and not for others is because it is a patch for a particular (hopefully common) form of inefficiency, but the patch only works if you have that efficiency or if the source of the inefficiency is the same.

I had a friend in graduate school that had a mental breakdown and lost most of her short-cuts. It was a really amazing thing to watch her make decisions from ground zero. I think she appreciated the experience because, to a large extent, she had the opportunity to pick and choose which new short-cuts to assimilate. As an intelligent adult that was more or less X-rational as well, she proceeded in a very systematic way. The result was a very efficient but somewhat strange decision making process.

Comment author: Nick_Tarleton 10 April 2009 06:49:34PM 3 points [-]

I think that a lot of advice represents "hacks" for getting around certain built-in inefficiencies. It would be enormously trouble-some (and arguably counter-productive) to rewrite your whole meta-code, but a patch is worth adding, if it works.

Strongly seconded. This also means that if one is unusually capable of self-rewriting, or at least deeper patching than usual, even common advice that works can be greatly suboptimal, or even harmful in closing off opportunities for greater growth.