Monday, June 25, 2007

Sim City

Researchers have found that it is possible to model the long term behavior of societies by focusing on a few factors, such as for example, the search for resources, rather than resorting to complex calculations of rational behavior. Joshua Epstein, a senior fellow at the Brookings Institution asserts that, "that certain sets of microspecifications are sufficient to generate the macro­phenomena of interest."

In plain language, Epstein suggests that a few key considerations motivate individuals and small groups of individuals; and that the collective interactions of these micro-actors determine the shape of a society. This eerie rediscovery of a "hidden hand" of individual motivation disturbed many social scientists, perhaps because it illustrated that social futures were much more sensitive to the motives of individuals and of small groups than to large scale social planning, in the manner of the Stalinist Five Year Plans

But the outcomes were not linearly determined by the choice of what motivation or values the micro-actors adopted, though that doesn't discourage the advocates of control. "If we're hoping, like Asimov, to predict the future, Epstein's models will disappoint. In fact, because his models give widely divergent results even when their agents are programmed with very simple rules, they indicate that predicting the future will never be possible. Still, Epstein's artificial societies do more to make plain the hidden mechanisms underlying social shifts--and their unexpected consequences--than any tool that social scientists have hitherto possessed. In the future, they and others like them could suggest how policymakers can engineer the sorts of small, cheap interventions that have large, beneficial results."

The "small, cheap interventions" that the MIT Technology Review article hopes for are likely to be elusive. If Epstein's research tells us anything it is that while affecting what makes people tick is the one of the most important variables of all, it is a variable that is surpassingly hard for central authority to manipulate in a unilateral way. Propaganda can't cut it. Central authority can only hope to deal interactively with individuals and small communities, which means that the dream of authority -- that vision of every social engineer from the beginning of time -- must always come hard up against the reality of freedom; mankind's curse and it's supreme blessing.

7 Comments:

Blogger Tarnsman said...

All which goes to show why the American system has achieved the success it has: individuals working in their self-interest furthering the interests of society.

6/25/2007 06:19:00 PM  
Blogger El Baboso said...

A few months ago the saying (attributed to Einstein) that "the definintion of insanity is doing the same thing over and over and hoping for different results" was very popular in the blogosphere. But as this study and chaos theory show, in a complex system you can start with identical initial conditions and come up with a different end state each time you run the process. Chaos theory also strips bare the utter hubris of "effects-based operations." MacNamara had his early version of EBO and found out that human beings have this stubborn way of not responding in the way he wanted to the stimulus he was providing. Sometimes you can only win by eliminating the free will of your opponent -- killing him or imprisoning him.

Free will is programmed into the very fabric of this universe. A chaotic infinity of quantum states collapses before our choices. Immediately, a new infinity of possibilities arises and we are forced to choose again.

6/25/2007 07:37:00 PM  
Blogger Marcus Aurelius said...

El Baboso,

Not certain about your claim as different results given identical initial conditions. The thing is though the conditions have to be identical and a tiny difference at the start can cause radically different outcomes.

Still just another example of how we do not need complicated nuanced rules (full of shades of gray) to create complicated & complex behavior.

Bill Whittle not too long ago started an essay (on Eject Eject Eject) based on the prisoner's dilemma and talked about the strategies various types of societies typically use when interacting with each other.

6/26/2007 06:14:00 AM  
Blogger Jrod said...

I wonder how Sim City treats the "Global Warming crisis?"

6/26/2007 07:26:00 AM  
Blogger Marty said...

Even Asimov's psychohistorians failed because they could not model the entire universe down to the quantum level (which, if one is to believe quantum theory, is probabilistic and may therefore not be even theoretically modellable with sufficient accuracy) and therefore did not predict the mutation of The Mule.

Which makes me wonder if Krugman even understood what he was reading...

6/27/2007 09:47:00 AM  
Blogger Red River said...

Its the Network, stupid!

6/27/2007 12:55:00 PM  
Blogger jj mollo said...

1) "Researchers have found that it is possible to model the long term behavior of societies by focusing on a few factors … "

2) "If we're hoping, like Asimov, to predict the future, Epstein's models will disappoint. In fact, because his models give widely divergent results even when their agents are programmed with very simple rules, they indicate that predicting the future will never be possible. …”

-------

Why do these statements seem contradictory? What he should be claiming is that models that include a few simple factors are able to simulate the superficial appearance of real-world phenomena. If a model is so completely chaotic that successive iterations yield dramatically different results, then it is of no use to us, cannot be distinguished from a model that is in error, and doesn't predict anything. Furthermore, in the real world we can predict events. Things tend to stay the same way for a long time. Many cultures are predictable within fairly narrow limits. The reason for this, and the reason these models are uselessly trivial, is that the real world is full of negative feedback loops that act dynamically to dampen wild fluctuations. Dangerous power centers tend to provoke countervailing alliances, for instance. Failures and crises will occur, but these sims don't help with that either.

Sadly, Asimov's vision is flawed as well for a related reason. Effective technologies will be copied. Psychohistory cannot be kept secret any more than nuclear technology could be. Contending groups will use the forecasts to further their own goals, just as they now use intelligence and analysis. Smart ones remain. Dumb ones disappear.

6/30/2007 07:28:00 PM  

Post a Comment

<< Home


Powered by Blogger