The upshot as I understand it is that some "Austrian Researchers" (watch out for those guys; they're nothing but trouble) wanted to create an artificial intelligence (AI) routine that was smart enough to play a computer strategy game, that also modeled human emotions. That is, their software would try to figure out when it was supposed to be happy, angry, cautious, apathetic, etc.; and then, it would behave in ways that a person who was happy, angry, cautious, apathetic, etc. would behave.
Then once they created their basic "human emotion" model, they used it to create several personalities. Now, a big part of our personality is expressed through our emotions. That is, some people are just happier most of the time; some are angrier most of the time, some are naturally cautious, or naturally exuberant, or naturally apathetic and nihilistic, or even just confused.
And even though we go back and forth between these moods, it takes different kinds of triggers for different people to effect a mood shift. So to describe all this, these researchers started drawing ideas from the field of Behavioral Psychology, in which one common Personality Model breaks down each person's personality into five variables:
- Conscientiousness--how self-disciplined and organized is the person?
- Agreeableness--how compassionate is the person toward others?
- Neuroticism--how easily does the person slip into negative emotions?
- Openness to experience--how easily does the person assimilate new experiences?
- Extraversion--how much does the person seek contact with others?
Guess which one won.
That's right, the most neurotic of the models performed best. It won all its scenarios against the AI opponent that comes built in to the computer game. Not only that, but it won each scenario significantly faster than the second-best of the AI players being tested, the aggressive one.
So, let's think about this for a second. The AI player following the neurotic strategy is designed to spend much of its time in a state of confusion; it's designed to make frequent mistakes about the availability of resources (the article mentions the timber resource as a good example). It's designed to have sudden, wide swings of simulated emotion, but on the whole to spend a majority of the time in a bad mood. It's designed to behave in a bi-polar fashion, switching its behavior suddenly and frequently.
And it wins.
A couple of thoughts present themselves:
- I think this counts as one more piece of evidence that Reason is highly overrated, something I've believed for a long time. Don't get me wrong; it has its time and place. And I'd much rather be a happy person than a neurotic one. But just because someone isn't guided by reason doesn't mean they won't succeed at what they're doing. Indeed, one's very lack of reason can be a significant factor in one's success. After all, your very randomness means that you will frequently change strategies that aren't working. And occasionally you will hit upon winning strategies by nothing more than pure chance. This means that in truly complex situations where the right answers are not easy to find, like in human societies--or wars--you are likely to land on a winning strategy sooner than one who is more methodical.
- Thus far, what we're describing is very similar in concept to Darwinian Evolution. Darwinian Evolution is driven by two phenomena: mutation, which introduces variety into the species; and natural selection, which weeds out the weaker of the varieties and permits the fitter ones to propagate into the future. Neurosis acts much like mutation, in that it causes the player to make frequent changes in strategy; and it also acts like natural selection, in that the player drops old strategies at the first sign of failure and replaces them with something new. This process can in fact yield quite successful strategies, without a whole lot of wisdom driving the process.
- If something causes people to be successful, it becomes commonplace. This is the whole idea behind natural selection: if being random, disorderly, bipolar, and undisciplined leads to a certain degree of success in, say, finding mates--and I do remember the high-school and college social scene, and this does happen--then one must expect randomness, disorderliness, bipolarity, and lack of discipline to become commonplace. If the Prima Donna gets what she wants, we all get more Prima Donnas.
- Ergo, our society develops a large number of messed up people, because at some level, it pays to be messed up.
It also seems to me that this dynamic works on much larger stages than interpersonal relations. One of the reasons there are so many messed up societies in the world, so many failed states, so many terrorist-supporting states, so many dictators nursing cults of personality, goes to the successes potentially brought by neurosis. After all, just consider: when a civilized nation is in conflict with a state run by a power-mad dictator, where do the sympathies of the "world community" typically lie? For the last several decades, the very fact that there was a conflict was laid at the door of the civilized nation. Power-mad dictators aren't expected to behave rationally, so it's considered incumbent upon the civilized nation to make accomodations to keep the peace.
(Boy, this turned into a depressing post. I didn't mean it to when I started it, really....)
I'm going to have to ponder what this all means at length. More on this when I'm through pondering....
No comments:
Post a Comment