In my explorations of intelligence, I learned something real about how the world works. Perhaps I am simply improving as an intellectual. Nerveless, I think there is a real correspondence between the actions and reactions of society and computation.
What is this thing we call society anyways? We can view it as a social network, with edges representing emotional closeness. But that seems to only reflect close connections correctly. Wars, duty, and other connections seem much broader. Then you start to realize that everything is a confusing mess that sociologists have been trying to parse through for years with limited success. As such, I think we should stay away from the specifics of how societies are formed. The specifics will be extremely useful for some purposes, like the creation of an ethical AI, but as long as we can’t grasp what societies do, the role of each root cause will be too complex for us to manage right now.
Instead, lets look at the incentives of a society, and try to derive the dynamics based on that. Unfortunately, societies don’t have obvious incentives, as we aren’t really sure what society even is. So instead, lets look at human incentives and how a society might protect and balance individual needs. We can cross-reference this with reality in order to try to prune out cases which might be true, but aren’t for modern society.
What do individuals look for in a society? Here are some cases I think form the core of society. But I have my biases and may be missing a bunch, or may be putting too much weight on some.
All of these factors conflict with one another, with many possible equilibria, leading to differentiation among societies.
Now that we see this list, perhaps you can see that it does not have to operate at the level of an individual. Families can have these views about each other in a tribe. Clans of tribes can have these view about each other in a confederation. Nations can have these ideas about super-nations. And this doesn’t have to tier in an orderly fashion either. The relationships between the nations of Europe and gypsy caravans followed these tradeoffs. The relationship between an explorer and a lost tribe can follow these as well.
All in all, what I am trying to say is the obvious thing that humans want things from each other, and find the dynamics that result. Note that I am not looking for equilibria, as standard economics suggests, because these systems are so slow to adjust, and the modern world is changing so rapidly that we will probably never be anywhere close to an equilibria in our lifetime.
So how does this help us reason about societies? Not much on its own. It requires tons of knowledge to be useful, at which you will already understand this. It becomes useful once you combine these ideas with how entities learn.
Some biologists have looked at society evolution similarly to a genetic evolution problem and are baffled by what they see. Genetic traits are pretty much indistinguishable from random traits just 3 levels of familial separation. So if we care about other people further than that, we are wasting our time. They will just become lazy and feel load off of us, disadvantaging us in favor of them. So societies cannot form in the same way that species does. One way to reason about this is that the societal entity must be smarter than genetic evolution.
So lets look at a much more powerful learning model: rolling ball gradient descent.
The idea here is that if we find that going in one direction is good, we keep on going in that direction until we find that going forward is more painful. And changes to our
a