My technical concern with this project would be the complexity. Errors accumulate within assumptions for models. A small error at a sensitive point in a model can seriously impact the outcome of a model run. Serious errors that generate ridiculous outcomes actually aren't so bad; the model is off and it appears to be off. What's worse is a not so serious error, one that leads to a plausible outcome from the model. In that case the model is off but it appears to be on. At high degrees of complexity the object interactions are so numerous that it is almost impossible to know if all of them went as the system designer intended.
...
SWS also replicates financial institutions, utilities, media outlets, and street corner shops. By applying theories of economics and human psychology, its developers believe they can predict how individuals and mobs will respond to various stressors.
...
- http://www.theregister.co.uk/2007/06/23/sentient_worlds/
Other links:
http://www.purdue.edu/acsl/abstract/march10_06.html
https://www.dmso.mil/public/
http://en.wikipedia.org/wiki/Synthetic_Environment_for_Analysis_and_Simulations
Here is an older (2001), general information article on the use of simulation games in training military personnel at all levels. The article mentioned a book about the history of video games called Joystick Nation that sounds like it may be worth a look.
No comments:
Post a Comment