My last examples have had individuals (or groups of them) as parts of the system, but the individuals themselves can also be modeled as systems with equilibrium. (if you care, this was first published by Powers in his 1973 book, Behavior, the control of perception).
To explain it I am going to bring the cruise control example back.
It isn't a perfect analogy but the differences are really interesting. For example, cars are really good at figuring out how fast they are going (low error) while people can be pretty awful at gauging how other's perceive them (high error) due to things like miscommunications, time delays, inability to read social queues, etc.
Another interesting comparison is that cars are really good at knowing exactly how to use their engine to get what they want to happen while people can be HORRIBLY BAD at knowing how their actions affect how other's perceive them.
So here we are with a speedometer that is hard to read, has a time delay, and the accelerometer and break seem to randomly swap positions or stop working. Social Anxiety. Anyways...
The biggest oversimplification in my eyes is that confining the concept of 'behavior' or 'perception' to a single number (Aka one dimensional) is a drastic oversimplification. HOWEVER, engineering control systems can be abstracted to as many numbers, in vector form, as you desire. You could represent behavior or perception as a 5 dimensional, 30 dimensional, or 185734589013 dimensional vector, if you wanted. Let's not try.
Powers goes on to attempt what I think is an equally unnecessary, and more subjective feat. Defining the ACTUAL CONTROL SYSTEMS WHICH WE ALL SUBCONCIOUSLY USE TO REGULATE BEHAVIOR. Apparently, there are 11 levels of control systems in human behavior. While the specifics of what level does what and how many there may be (nick's hot take) a bit of a fool's errand, the concept of lower and higher level control systems is really valuable.
In this example and in general,
Lower level control systems are more concrete things like the tires and the ground. As you go 'up' in levels, you increase in levels of abstraction to things like how fast you think you should go, what type of gas or oil you use, what car you own, which road you are on, the weather conditions, or the government adjusting speed limits. All of these affect your car's speed. The top-most levels control system is the most abstract in the system, but receives much steadier reference signals. Where you are driving to. Prioritizing safety.
Only the lowest level control system (our words and actions) sends an output to the environment and only the highest level system receives a steady reference signal (things our values and past experiences)
Higher level control systems determine the ultimate goal and lower level ones operate on smaller domains determined by the high level system and the environment.
If we, god forbid, implemented this in a game like WoW and turned it up 'to eleven' as Powers does, it could look like this:
I don't think all of that needs to be automated. Some of it sounds fun to try to automate, but some of it sounds fun and expressive to do. Like as a job. Don't automate the art jobs please.
This is relevant to the system at a whole because by understanding how to model an individual, we can understand how they function as a part of some larger group, and then how groups function.
I'm not particularly interested in mathematically modeling how an individual behaves, but I think an understanding of how individuals behave on average is more interesting, feasible, and useful for things like social system design.
Everyone behaves in response to their environment. This isn't just an adult thing, or a normal human thing, or an animal thing. Plants do this. Bacteria do this.
🐢
🐢
🐢