In a system of mechanistic causes and effects, meaningful and complex states don’t just come about, and will fall apart quickly as things continue to happen. This is entropy, the tendency of a system to devolve into more likely, more random, more chaotic states. (Wikipedia)

Attractive Forces

Except when attractive forces are involved. In such a system the more likely state is everything clumping together. In our physical universe, matter is attracted to other matter under the force of gravity. This causes a pattern of globes with discs1 at various scales: galaxies; star systems; even planets can have rings2. Another example is crystals, which are beautiful arrangements of molecules or atoms coming together under electromagnetism in such a way that the underlying structure is still visible at the macro level.


When some order appears in the system, energy flows can become restricted, that can have wave like effects, or channel like effects, or various other effects. This and the attractive forces, means our universe is far from boring and chaotic all by itself.

For example, under enormous pressure, stars start shining from nuclear fusion, beaming huge quantities of energy into space. On a planet like ours, this causes day and night cycles, hot and cold, seasonal effects, water cycle effects. Rivers will form, carve out riverbeds and get entrenched in their flow. Hot magma can deposit at the surface with a whole range of interesting effects.

Learning Systems

But there is one effect that can truly go against the grain of entropy and create unlikely states far beyond rivers or crystals: learning systems. Systems that can adjust their responses to the situation. That by negative feedback learn the behavior wasn’t effective and try to update, so that future responses to similar situations are more useful.

We as intelligent beings do this by studying, reasoning and remembering knowledge. But we also do this through experience, by training our unconscious reflexes and intuitions. You cannot learn football in a classroom or from a book, you have to train. Most animals can also learn in this training way, even insects and mollusks and many others3. And there are many algorithms that we can use to train a computer, to do things like recognizing speech or recognizing cats on youtube. (Wikipedia)


If through various self-organizational effects, some thing can come about that can shape other things into a copy of itself, that also creates a learning system: things that can do this will continue making copies; and as long as conditions are good, grow exponentially. Things that do this the best will grow the most, pushing out others that are less effective at making copies or don’t make copies at all — an automatic self-selection effect.

If it is possible that the copies aren’t always identical, and that some versions might be better at copying, the self-selection effect will cause these to become the majority. Once a better version appears on the scene, it is unlikely to disappear, unless even better versions overtake it. This process goes faster when the environment changes, or when there is competition for space or for resources. Or when more complex interactions appear, like symbiosis, or predator-prey relationships. (See abiogenesis and evolution)

Goals and Purpose

Because entropy destroys complex configurations of matter, the only complex configurations that can stay around are those that work towards the goal of not disappearing4. This is at the root of all meaningful and complex states in our universe.

This process caused brains to appear that can be trained, and even the human brain, the most intelligent brain to date. What brains are for, is to try and predict the future5; to be able to react to danger before it is to late; or recognize food and other useful situations, and react appropriately.

Brains that can recognize situations can have goals that go beyond not disappearing, like avoid pain, avoid hunger, or try to be safe, try to mate.

Brains that are intelligent, that can imagine a whole range of possible and far away futures, can set even higher level goals for itself. And even, to some extent, override goals set by the environment or the body. For example, collect and store food for the winter, neither the environment nor the body will reward this behavior. But knowing you won’t starve in the future is rewarding.

Free Will

We are at the root of so much anti-entropic possibilities. Here because we learn from experience and through our intelligence. That learning shaped who we are. While the causal chain of the universe has no choice, mechanistic causes and effects must follow our brain chemistry, must follow where our thoughts lead it, must follow as we learn.

From the perspective of the universe, our path was set the moment we were born, we are not free from the causal chain. This leads some to say free will is an illusion6. But from the same perspective it is our learning that is at the root of so much meaningful change, creating so many unique and unlikely things, against the grain of entropy and far beyond natural things. Your decisions, what you learn, what you choose, matters a lot for how the future around you will unfold.

From our perspective, the universe has no choice, it is not free, it must allow our thoughts to flow, it cannot limit our imaginations, it cannot influence our choices. And when we do take action, the same rules continue to apply unchanged, it is the universe that is not free.

Is this enough freedom for you to call it free will?

By analogy, from our perspective, the ground is stable and standing still under our feet. But from a frame of reference outside of our galaxy, we are on a spinning ball, rotating at enormous speed around a sun, itself rotating around the galactic center at a speed close to a million kilometers an hour. Stable ground is as much an illusion as free will. Frame of reference is everything.

  1. A disc appears as matter is sucked in, because just like in your bathtub, a twister appears. But unlike your bathtub, in space there is no friction, so the twister never loses its energy. Matter in the twister can only get closer to the center by moving towards the plane of rotation, creating a disc. Also any pre-existing angular momentum is amplified, and frame dragging plays a role in the largest of structures. (Wikipedia

  2. Planetary discs are often there from another effect, gravitational sheering effects on moon like bodies can tear them to pieces, creating a cloud of debris in similar orbits. (Wikipedia

  3. Can insects learn 

  4. Notice this is not a claim anything or anyone had some kind of intention, or made a choice, to adopt this as a goal. Instead, many systems will have many different effects, but only systems which have an effect that they stay around in face of entropy actually do. And that such a system can be correctly described as working towards the goal of staying around. 

  5. Do our dynamic brains predict the world? 

  6. The illusion of free will