Todd Hylton Is the universe a product of thermodynamic evolution? TEDxSanDiego 2017
So in 1967 Richard Brautigan published a poem, and it was titled All Watched Over By Machines of Loving Grace. And he suggested a future in which our technologies, our machines, had such an understanding that they would care for us or relieve us from the struggles of our everyday life. So today we are again fascinated by a future with machines. Things like artificially intelligent computers, and self-driving cars, and robots.
It seems to me though, that hope for machines that would take care of us has been in large part taken over by fear that machines might replace us or even destroy us. So in this talk I want to address these hopes and fears for our technologies. But to do that I’ve got to tackle some really tough questions, questions like, is the world a machine, and how do things come to exist, and can we build technologies that understand the world all by themselves.
Now I’ve spent the last ten years rearranging my brain trying to answer these questions, and I’m going to spend the next ten minutes trying to rearrange yours. So hang on. So at first I’m going to talk about what it is to be a machine. So I think there are two things that make machines different from us, and from what we typically think of as the natural world.
The first thing is that machines are essentially the sum of their parts. And by that I mean we can take machines apart into pieces, but we can also put the pieces back together and recover the entire machine. This also means that we can replace broken parts, and that we can make copies of machines by putting together similar kinds of parts.
Now this is very much unlike us. We are indeed made of many parts but we can’t simply remove and replace them and we can’t make copies of ourselves either. A car losing a headlight, and a person losing an eye aren’t the same thing. Now the second difference is that machines are connected to the world through us. My car drives because I drive it. My phone makes a call because I call somebody.
Everything that machines know about the world, they know because we have built it or designed it into them. Again, this is very much unlike us. We know about the world because we’re deeply embedded in it, and we change as it changes, and we could say the same thing about all living things. My dog knows all about his world without any help from me.
So I think in everyday life most of us realize that we’re not machines, right? Real life is just too complicated and too unpredictable to be a machine, as almost any family gathering would instantly confirm. And we actually get into trouble when we try to describe the world, as a kind of machine. So we get into trouble. So here’s a story to sort of illustrate what I mean. So suppose that you and I want to go out and we want to build a forest on a barren plot of land. And so we might look at a real forest and see there are trees. So then we might make things that look like trees out of plastic and put them in the ground with concrete.
And I might say, hey you know that forest has a stream. So then we might dig a ditch and pump water through the ditch. And you might decide that what we really need is to make leaves out of fabric and attach them to the trees with actuators that release them in the fall. And of course this is a terrible idea. It’s a ridiculous idea right. Everybody knows that if you want to build a forest you need seeds and water and air and sunlight and time for a forest to grow. So the point is that a forest is an evolved entity that is intimately connected with its own environment, and it cannot be assembled from a collection of parts. So a forest isn’t a machine.
So given this simple story though, you might be surprised to hear how much we want to describe the world as a machine, and as an example the pinnacle of scientific achievement, is a description of some part of the universe as a simple mechanistic formula. Something like Newton’s famous law that says force equals mass times acceleration.
You see the great thing about machines is that, because they’re disconnected from the world, and because they’re essentially a collection of parts, they’re simple, and it’s easy to predict what they’re going to do. But if we take this love of simple things to its logical extreme we come to the idea of a universe that’s completely mechanistic.
And when we do that, we also arrive at the conclusion that we need a supernatural entity to create it, just like a machine needs an engineer, a mechanistic universe needs a creator. Now in my experience this idea that the world some kind of machine is actually really common, even among scientists. So curiously I find that like some religious fundamentalists some scientists are actually creationists, although maybe they don’t see it that way.
But the idea that the universe is a machine is not what science actually tells us. If the universe were a machine then science would come up with simple formulas that describe everything, even large complicated things. But this isn’t what science finds. What we find are enormous numbers of specialized descriptions of the various parts of the universe that aren’t simple.
We only find simple descriptions when we look at simple things, and this is how we get things like Newton’s laws. So to say it differently, science shows that we can take the universe apart into pieces, and describe the pieces, but we can’t put the pieces back together into a simple understanding of what the universe is. So science shows that the universe isn’t a machine, but if the universe isn’t the machine, then what is it?
So the house that I grew up in, in Virginia, had rain gutters and downspouts to concrete blocks on the ground that channel water away from the house. And when I was a kid I used to run around and turn over these concrete blocks to see what was underneath. And I remember this really well because there were all these bugs underneath there, you know, really weird looking stuff, like centipedes and pill bugs and worms and their larva and carcasses.
And I was always fascinated because the bugs were doing things, and I thought how do they know what to do? Why are they there? Why is there a world under this concrete block, and who could possibly care enough about them to create them?
And though I couldn’t really say so, say why at the time, you know these bugs under the mud, in the mud, really bugged me, and they bugged me because I came to understand, that I feared that I might not be that much different. Right?
How did I come to exist. How do I know what I know, why do I do what I do. And more recently, as I address the problems of, you know, can we build technologies that know about the world all by themselves the same questions come up. Right? How should such a technology come to exist. How should these technologies know what they know or know what to do.
So, in a mechanistic universe the answers to these questions are simple but they’re not very satisfying. You know the answer is that the Creator made it this way, and this mechanistic universe is evidently incorrect, both in our personal experience, and sort of in the broad efforts of science, but I think we cling to it nonetheless, at least in part because we can’t explain how things come to exist.
So I would like to offer a non-mechanistic, anti-supernatural, evolutionary worldview of how things come to exist. I propose that the universe is uncaused, incomplete in and of itself, and that it began with a featureless singularity of pure potential. A big bang. I propose that the universe derives from the pouring out of this potential into the creation, the evolution, and the existence, of everything in the universe.
I propose that this thermodynamic evolution is not a question of mechanization, it’s a question of self-consistency. I propose that causation, the idea that one thing follows another in time, is the result of evolution and not the other way around.
I propose that the universe fits together because it derives from a common origin and evolves under a common principle. I propose that every aspect of our lives reflects this thermodynamic evolution; our birth, our development, our thinking, our interaction with others, or interaction with the environment, our technologies, our evolution as a species, and our death.
So simply put ,I proposed that existence derives from thermodynamic evolution. Now some of you may have heard of the second law of thermodynamics, the idea that the entropy of the universe, which has to do with the distribution and disorder of energy, is always increasing.
So the idea of thermodynamic evolution is exactly this, except it adds the requirement that an organization must emerge for that to happen. So in everyday language, thermodynamic evolution is simply the idea that organizations spontaneously emerge in order to use energy and the environment. For example, wood is a kind of concentrated chemical energy that when we burn it with oxygen releases heat broadly into the environment.
But in order for that to happen, an organization has to emerge, a fire in this case, that connects the oxygen to the wood at high temperature. Now the oxygen and the wood don’t cause the fire in a mechanistic sense, only in the sense that they create a thermodynamic potential such that a fire can emerge if the conditions are right.
As another example, I am made up of atoms, but the atoms of which I’m made don’t cause my existence. Energy and the environment, in a complex process of thermodynamic evolution, enabled the atoms that are me to come together, and the same can be said of human organizations like businesses. So businesses are collections of people and assets, but these people and assets aren’t a machine that causes the business to exist.
Businesses evolve in a complex environment and extract energy from the environment to do that. Now the reason that the people and the assets fit in the business, the reason that the atoms in my body fit together, the reason that trees fit within the forest, is that they evolved together. So things that don’t evolve together, don’t fit together. That’s where expressions like fish out of water come from.
So you would think we would have figured this out by now, but we haven’t figured it out yet. If I look back on my childhood, you know the bugs in the mud, I would say that, you know, no computer, no phone, no robot, could survive under a concrete block or anywhere in the real world on its own. The problem with our technologies today is that they can’t evolve to fit into their own environments. They only evolve if we do it for them.
That’s why you get constant software updates on your phones and your computers. So the net of all this is that, if our machines are doing good or bad things for us today, it’s our fault, all right, we’re the ones doing the evolution for them. I think however that we’re on the cusp of being able to build electronic systems that can evolve.
So the interesting thing about electronic systems today is that the basic elements, things like transistors, have important things in common with the basic elements of biochemistry, things like proteins, so transistors and proteins are roughly the same size. They can both be configured to form complex networks, and the energies and the temperatures that matter in the formation of those networks are comparable.
And for these reasons I think that we can build electronic systems that spontaneously evolve complex networks in response to electrical potentials in their environment. Now we don’t know how to do this yet, and we’re really not even trying, and I think the reason we’re not trying is we’re stuck with a largely unrecognized mechanistic worldview that prevents us from knowing what it is we should try to do. So what we do instead is we build bigger machines, and we program them.
Someday we’re going to get this figured out, we are going to build systems, I call these things thermodynamic computers by the way, and someday we’re going to figure this out, we’re going to figure out how to do this, and we’re going to build technologies that aren’t machines, that can evolve in the world on their own. And when we do that they’re going to grow up in environments that we curate. They’re going to be more like our kids than our computers.
Which should be interesting, actually. So it may be possible that they could learn enough about us to actually care for us as individuals. And so, perhaps this vision that Brautigan had five decades ago of technologies of Loving Grace is actually possible.