RE: this thread: http://www.reddit.com/r/PurplePillDebate/comments/2jfnf4/compulsive_systematizing_vs_compulsive_empathizing/

Read this: https://newsoffice.mit.edu/2014/autism-disorder-prediction-anxiety-1007

Summary: there is a (IMHO decent) hypothesis that BP tends towards compulsive empathizing and RP tends towards compulsive systematizing, having Asperger-like, slightly autism-spectrum characteristics. (Note: undiagnosed, but I suspect I may be a bit on the spectrum myself, so don't take this as an way to undermine RP. However I scored low on the systematizing test in that thread because that measured whether I like to, not whether I can. I can, that is my job (ERP consultant, in older times that was called system analyst), but feel little desire to do so as a hobby. This may be part of the reason why I like to lean blue.)

Anyway, this article is basically saying the spectrum may be about low prediction skills, be that human interaction or where a ball will land. Compulsive systematizing, rigid structures, repetitive behavior are ways to create a more predictable personal environment.

However. The most predictable thing is when nothing happens, or when something truly simple, mechanical, repetitive happens, like with mechanical toys. Yet, most Aspies tend to like things like computer programming, which is not simple at all. Is a typical code predictable? Yes and no. Certainly it takes a lot of learning, practice, experience and work to predict the reason of a more complex bug. However, this is a different kind of prediction than figuring out where a thrown ball will land. It is more recipe-like, more rigid, more logical. This is entirely conscious, verbal, intellectual, not instinctive. This kind of calculation seems to be happen in the very conscious layers of the brain, actually the in roughly the same layer which IQ tests measure. (Not only Aspies tend to have high IQ, they even tend to like the kinds of tasks IQ tests use.)

My hypothesis would be that a person really cannot live functionally without predictional abilities. So the same way how blind people have keener hearing, Aspies tend to use conscious, intellectual processes for prediction instead of the dysfunctional instictive skills, which exercises and grows the IQ-muscle, the logic-muscle. How to piss off an Aspie-type programmer? Put him into a situation where he must accept something illogical, like a contradictory requirement that is nevertheless accepted because the people who pay the bills want it so.

One interesting parallel: RP type overgeneralization (AWALT) seems to map to me a typical attitude of computer programming blogs (that I find similarly annoying): that once they find something that works for them they insist that absolutely everybody ever should be doing that all the bloody time. For example, that ALL programs should be covered by unit tests, NO exceptions. So the tendency to form these rigid ideologies. And I have similarly frustrating debates there, trying to tell the authors there are different fields of programming with different conditions, like e.g. when theoretically every routine could look up any field in a 100GB database and change its behavior accordingly, there is no way you can unit-test that. And it similarly tends to fall on deaf ears.

I should also say that I was similarly rigid when I was young, up to 25 roughly. I took an interest in buddhist meditation, did it for years and it seems to have dissolved this subset of my Asperger, this strong desire that reality should comform to rigid rules. Formerly I talked in black and white terms, X is true, Y is false, like that. Later my thinking got more perspective-aware and uncertainty-aware (probablistic, even Bayesian), now I am more like "It seems to be X tends to be more often true than false".

BTW the lesswrong.com community tends to be both Aspie and quite probabilistic. I wonder if it has any overlap with RP? My hunch would be not much.

It's a bit late around here, and I can sense I am rambling. Anyway, I hope this could be useful for someone.