My friend Charlie Yokomoto, a Professor Electrical and Computer Engineering, and a savvy observer of the political scene, has read this blog and offered the following thoughts on the difficulties of managing complex systems. Maybe we human beings aren't smart enough to deal with these devilishly complicated systems (think financial markets!) that we have created!

Here are some thoughts in engineering terms about the difficulty of decision making in modern society. One of the things that engineers do is to model a system with mathematics so that they can predict behavior to different inputs and find ways to control the system. The system can be a rocket to the moon, a car's transmission, a car's engine, etc. These are fairly basic systems to model because they can be treated as systems whose parameters don't change, or if they change, they change in easily describable ways.

When the system is more complex, like the stock market, the human mind, the atmosphere, a tornado, etc, the modeling becomes far more complicated. When society was much simpler (all white, mostly all middle class, mostly traditional families, mostly church going, low crime, low poverty), then the modeling becomes simpler, and making rules to control the society was simpler than today.

Now for complex systems. A semi-trailer truck sliding on an icy road can also be modeled, but it is more difficult. Mathematics profs are now modeling economic systems, where parameters are always changing. Probabilistic methods have to be used, and time varying dynamic equations have to be used. The atmosphere has been modeled with time varying parameters for years. These equations are more difficult to solve, making it more difficult to use them to predict things, and to control it.

OK, now about controls. Then a system can be accurately modeled, then mathematical methods can be used to determine if the system is observable from the outside (can you observe all of the changes, or will some changes be hidden from the observer?). Methods can be used to determine if the system is controllable (are there ways to push the system in a direction you want it to go?) Sometimes, you find out that you can't--the system will go where it wants to go.

As systems become more complex, you then find out that your mathematics is not equal to the task of modeling it--you need more complexity. Example. If you have a groups of points on a graph, and they all line up in a straight line, you can easily model it with a straight line. If there is a slight curvature, you can try modeling it with a second order equation. But if it is fluctuating wildly, then you need a very high order equation to model it.

I believe that society is getting to the point where it is getting so complex that the human brain cannot intuitively, through experience, fully understand develop a heuristic model of society--how creation of jobs, curbing crime, helping the needy, paying for roads and schools, keeping the food supply safe, etc.--can be done using the same decision making processes that were used successfully when the system was much simpler (fewer parameters, slower changing, and more homogeneous). If the system become more highly complex at rates faster than the human mind becomes smarter (not necessarily in IQ, but in terms of ability, tools, and know-how), how can they solve problems in a system whose complexity if more that their brains can handle?

Take crime. If you have 100 cops and 1 person breaking the law, they can handle it. If you have two people breaking the law, they can handle it. What if there are 10 people breaking the law? Twenty? At what point, will the system of people breaking the law become too complex for 100 cops to handle? Some say that you can't have a perfect law enforcement system and that you have to tolerate a certain level of crime going on at any time. Well, people in charge were ok with that as long as the crime was kept in particular neighborhoods.

Anyway, that's how an engineer looks a society, which is just another system--a very complex one at that, that becomes more and more difficult to model and influenced as the complexity increases, possibly to levels that cannot be understood or influence with legislation and rules.

## No comments:

Post a Comment