Saturday, March 22, 2014

Complexity - 1

Living things are perhaps, at root, energy conversion machines.

Think about if you were creating a universe and you had two options: one to design everything from day one, the other to put in places machines to do the work for you. Looking at ourselves as humans today, we know that the second idea would appeal to us. Making and using tools and machines comes naturally to us. Tools which amplify our abilities and extend the range of the possible. 

But we know that machines consume some form of energy. So if we had a lot of things to build we would come up with two things: First the machines, and second a whole lot of energy to run them. And we'd set things up in a way that the machines can convert this energy into useful work and in the process transform the space in which they exist.

In a way, this is entirely what happens. All living things basically absorb energy and make an impact on their surroundings by transforming this energy into work. Plants do it, animals do it, and humans do it perhaps most visibly. But the concept remains the same.

Of course if you were thinking very long term, you would do one more thing in addition to having energy and machines. You'd think about learning. As the machines create one set of outputs, the next task would be more complex. They would start with the easy ones and make their way upwards towards the more difficult ones. When we look around this is exactly what we see, we build more and more complex things, transform energy in more and more useful ways as we learn more. So in addition to having lots of energy to work with, the machines should have the ability to learn from their own work and adapt. 

This is how complexity works. What seems remarkably complex today often arises from simpler rules working together and interacting over increasing scales of time and space. Complexity thus emerges as a consequence. What seems complex today will seem trivial in a hundred years - just like what appeared complex a thousand years ago appears to be trivial today.

Three elements are therefore important: Lots of energy, machines to transform this energy into work, and a feedback loop which enables these machines to learn and adapt. Leave these three elements alone and results would likely be spectacular.

No comments:

Post a Comment