Democracy, Entropy and Generative Adversarial Networks
Entropy dictates everything around us changes. No way around it.
What this means for everyone is that everyone needs to be agile, capable, as an organism or a system, of adjusting to changing conditions, changing contexts.
The direction of time is towards where there is more information. This is what Shannon’s Law tells everyone about entropy.
99.9% of eukaryotes reproduce sexually. Reason is that it is more advantageous for adaptation to have access to more information. Male and female genes together constitute more information. More information, more capacity to adjust to where events go.
These are all, of course theories. Generalizations. “Entropy dictates everything around us changes.” “No way around it.” “Everyone needs to be agile, capable of adjusting to changing conditions.” “The direction of time is towards where there is more information.” Shannon’s Law.
Taken as theories, none of these generalizations can be said to be true. Only reliable for predicting.
Science is, one way of looking at it, a body of generalizations connected to one another as a system. Scientific generalizations themselves are products of the use of the scientific method. The scientific method itself is considered a more systematic, more methodical, form of induction, meaning a systematic form of learning from experience. From the body of experience itself of the scientific community since the advent of science in the 1500s, everyone has learned that generalizations produced across time end up either being revised in form or value, or even rejected in favor of better ones — better meaning they provide more reliable explanation and, more importantly, prediction. Ergo, scientific generalizations are never considered “truths.” All are expected through time to either be revised or rejected in favor of better ones.
Even science itself has to contend with entropy, adapting theories to demands of the movement of time.
Generative Adversarial Networks (GANs) are effective in many ways. The assumption with the development of GANs is that more information is better. Meaning more information regarding some objective.
With GANs, the objective is attained through competition among systems from which the objectives can attained in unique ways. For example, take some competition among a sundry of “mind models” competing for making some virtual machine attain mobility, capable of walking like humans do. The more competing models, the more information, the better. Some lose, some win. In the end, virtual mobiles end up efficiently moving about in entirely surprising ways.
GANs, of course, assume natural selection. Some systems lose and die. Some systems win and persist in some form of existence. In the world of GANs, systems run by Emperors with New Clothes become extinct very quickly.
Entropy entraps the universe in a way that a Laplacian determinism loses out in a world where systems have more choices, more information. Laplacian world is like the world of Emperors with New Clothes — one has no choice but to follow the emperors wish. Laplace’s demon cannot stand competition against Maxwell’s demon who can gain control over his destiny, thanks to information. More information wins. Laplace’s demon is a pauper in this sense.
Entropy necessitates freedom, more access to information, information that Maxwell’s demon is very good at using as a tool for control. More freedom, more information, more control.
Social and political systems are no different from organisms. Same with cultures. All systems compete. Those that can harness more information, gain more control. Adapt better. Survive better.
There is more value to democracy when viewed from the point of view of Shannon’s Law, from the point of view of the demands of entropy. The world of Information Communication Technology (ICT) is replete with works making explicit the demands of entropy. It is only a matter of time that ICT whiz kids at large are going to understand this. They are going to see for themselves that freedom, democracy, entropy and Generative Adversarial Networks all are members of the same set.
They can be expected to genuinely value democracy. More than my generation.
How entropy works. New waves push old ones.
(Photograph by Joe Galvez. Colorized by Prince Javier/Bridge360AI)