The power of computers is advancing at a rapid pace. Eventually, will we have the power to begin computer simulations of whole human populations? What would this mean economically if we were able to run these types of simulations? Could simulations of economic and political policies first be run on a computer before being enacted in the real world? Would this result in an optimization of economic activity and an ever greater increase in gdp per capita? How does one successfully model human behavior to a degree that is both accurate yet computationally efficient?



The human brain is one of the most complex organs known to man. To successfully model human populations it is vital that the computer models can successfully account for how an individuals brain would respond in specific situations in a sensory environment. Perhaps in the future, computer simulations that model entire human brains will allow even more realistic virtual simulations. However, this would require too much computational power that is not currently available. The Pentagon currently wants to "realistically replicate human behavior and frailties" to improve cyberwarfare tactics. The philosopher Nick Bostrom has argued that we may already be living in a computer simulation being run by an advanced civilization. This seems quite far fetched and I find it unlikely that we are in a simulation being run by other beings. However the increasing in computing power available to us may mean that we will soon be able to run these advanced computer simulations ourselves.

I think it could potentially be enormously beneficial to society to accurately model human populations. I would suggest that reliably simulating this aspect of general intelligence (along with possibly several other psychological variables that are too complex to go into in this post) might result in a reasonably accurate simulation. Now the simulation would currently be constrained by computational capacity. It could however allow a future society to run multiple simulations with various economic policies to find the one which lead to the best outcome for every person. A simulation is only as good as it reliably models human behavior. So they might have to run multiple simulations with different variables until a model that worked reasonably well was chosen.

A computer simulation would have the same problems that economics is plagued with currently. How do you measure economic success? Some people place more value on gdp per capita, while others might place more value on things like universal healthcare. I think it would be difficult for people to agree on what was the best measure of success for a population. There is also the issue of control. A person who successfully creates a computer simulation that can reasonably model human populations might have a considerable advantage economically over other people that did not have the model.