Can human interactions be predicted through computation?
In “human complex systems,” or HCS, the answer is yes and no. Likely interactions can be determined, but the future cannot be predicted.
Human complex systems are basically a way of thinking about human interaction through artificial realities. Think “The Sims,” but with broader implications for social science.
The virtual worlds help us understand how humans behave in patterns through simulations.
Challenging assumptions is in some ways the basis of these simulations.
“What I think the real value is, is the fact that (using human complex systems) forces you to take ideas about society and about human interaction and make them explicit,” said Nicholas Gessler, an executive director of the HCS program at UCLA.
These simulations can be used to recreate cultural differences, such as soldiers in foreign countries or spatial movement in evacuation processes, or the way humans move throughout space in buildings we use everyday.
Human interactions, whether they are people in stairways or international relations, cannot be predicted exactly, as far as the most likely outcomes are concerned. People are unpredictable after all.
It’s like imagining a moving car on a freeway; thinking of all the moving parts at once is nearly impossible, Gessler said.
What can we learn about real life from modeling these types of systems?
An effective use of these simulations is evacuation procedures. Simulated humans are given certain behaviors that mimic those of masses leaving buildings. What can be found through simulation is that if a cylindrical obstruction is placed in a door, the flow and speed of movement increases, said Christian Bak, a history student who is minoring in human complex systems.
These equations become more interesting and more complex when cultural cues, signals and behaviors are added into the equations.
Human complex systems simulations “take cultural variables and cultural cues from individual agents and have them affect other agents,” said Alex Crown, academic and research director of the Human Complex Systems Society, a student group.
These types of multi-agent systems can be executed in real life, too.
Dario Nardi, who is an executive director of HCS, noticed a lot of chaos in the stairwell and decided to do a multi-agent, real-time experiment in his Formal Modeling and Simulation in the Social Sciences class to organize the flow in the stairwells. The students of this class started following structured paths at certain times of the morning ““ one passageway was up, the other, down. Within three days the entire building followed suit at these times, Crown said.
A large group of people can regulate a space through unspoken rules.
The simulations can be more game-oriented or more like social science, because either way the students creating them “are learning the same processes and the same functionality that they can use in serious science simulations,” Gessler said.
These factors can be cultural, behavioral, spatial, environmental or political.
In most cases a visualization is what is put out through these simulations. These visualizations are not seen as a place for enforcing assumed qualities, but are a space for challenging assumptions about human behavior for clarity in complex ideas in social science.
“People complain about scientific reductionism. This is sort of the opposite, this is a synthesis: Take a problem and reduce it to the lowest common denominator and assume that all the parts in operate the same way,” Gessler said. “In multi-agent simulation, you try to describe the parts and from that you synthesize something bigger, which quite often doesn’t behave the way you intuitively think it should, and that’s fun.”
The visualizations allow the perspective of the viewer to encompass all the actions and motives of moving agents.
“By teaching this, hopefully some students may be interested in building this sort of thing. Hopefully they will become more critical of ideas that are not tested in this way and also critical of simulations that policy-makers and politicians may be trying to sell,” Gessler said.
He said he hopes that anyone familiar with the idea of complex simulations will challenge the assumptions put forth by simulations, and want to look at the code to understand the underlying principles at work.
I hope the ideas in the human complex systems will allow a certain amount of transparency in our use of not only media, but the code working the devices we use.
Simulations can give insight for anything from political policy to which stairwell to use.
Social principles and individual behaviors can be explored through computing, and the very assumptions that are used to create these simulations can be challenged again and again by changing a few lines of code.