Friday, February 08, 2008

Have you any clients from Saudi Arabia named bin Laden?

Decartes wondered if we could put the world in equation; clearly this is not an uncommon goal. The British have an excellent weather prediction system--they have simulate the atmospheric conditions over all of Great Britain, in one big fluid dynamics petri dish. As a result, they can tell when the rain will hit your part of London, almost down to the minute: just feed in the state of the world as it is currently, run the simulation faster than real time, and there you go. It’s your window on the future. The world remade in equation.


This, of course, is of interest to any game coder. Because that’s all a game is, really: the world remade. We make everything, from the dirt to the sun--you can’t buy it, you can’t pick it up and put in it frame, you can shift the camera to see it. You have to make it, and you do that with it’s primitive component types. Coding is a process of giving things names; games coders seek the true name of the world.


So, we want to see the world remade in equation. So do cognitive science researchers, and AI researchers. Map the brain; and make it anew in logic. Find the weights, connect the neurons, map out the algorithms.


But if we can make the brain with numbers, where’s the magic? Lost in the quantum uncertainty, perhaps--but let’s say it’s not. Let’s say, human behaviour is equation. Let’s say, the product of human behaviour is equation. Let’s say, any society is like physics: we can model it accurately.


But if that’s the case; we can model the behaviour of society so long as we know the correct input to the algorithm. Sample the collective conscious, feed it in, and run it faster than real time: we’ll see when the spreading Obamaism hits Washington.


If we can do it for the weather, then why does it feel wrong to say we might be able to do it with people, also? Perhaps it’s because it feels like capitalism, and we hate capitalism.


If we assume that all agents are equal, and have equally free options, and work for selfish but mutual beneficial gain: then we can build an economic model of the world. We’re still working on that model, the study of economics is slower than physics, but it’s coming. We’ll feed in an easily sampled initial state, and find out the future.


But; if we assume that all agents aren’t equal, and do not always work for selfish gain, then it won’t work, right? We can’t model the individual will of people; we can’t model diversity and ingenuity, right?


Of course we can. Why not? Historians arrive are predictions and conclusions through rational models. Sociology does, also. The most primitive forms of statistics and social models are in daily use. We can model large scale social behaviour, even when our subjects are active, involved and intelligent. And what’s more; we can do it with economics. Because economics, above all, is the quantification of society, our way of counting the uncountable.

0 Responses: