28 June 2012

Sterling on Turing & gender

Author Bruce Sterling almost never fails to come up with interesting takes on a variety of subjects.  He recently gave a talk on the occasion of the 100 year anniversary of Alan Turing's birth, dealing with the Turing test and more broadly about cognition vs. computation.  Sterling points out that Turing's original description of the imitation test is like this:  "In the original Turing imitation game, you’ve got three entities: a judge, a woman, and a machine pretending to be a woman."  Sterling spins out some ideas around the role of gender in consciousness, in AI, etc.  Definitely worth reading in full, but here are a couple passages of interest to this blog.

You could argue that “masculinity” has nothing to do with "intelligence." I might even agree with you, but if my masculinity isn’t an aspect of my so-called intelligence, what is it?

Mathematics may be sexless, but do we really believe that cognition is some quality we have that is strictly divorced from gender? How can you properly claim that you understand how human brains work, if you can’t create a system that expresses a female sexual identity? Because billions of brains do that every day, and it’s not rare, because women are the majority gender. Where is that aspect of human intelligence supposed to be hiding? Is femininity non-algorithmic? Is femininity a Turing non-computable problem?
---
Sexuality is eons older than intelligence. We’re not abstract mathematical systems somehow burdened by gender. We are living entities produced by sexual means. Those are the facts of life.

We don’t yet know how cognition works. It wouldn’t surprise me to learn that sexual hormones, such as estrogen and testosterone, are fundamental to cognition and even to conscious self-awareness. We should have a spirit of humble inquiry toward cognition. We know far more about it than we did when we invented body-mind duality, but it’s a large, dark area.

No comments: