Every observable phenomenon in the universe can be modeled by a neural network. The universe may be a neural network.
A paper written by a professor of physics at the University of Minnesota Duluth was published on the arXiv pre-print server. The paper was published today by Futurism and it was covered by an interview with Vanchurin.
The big idea.
According to the paper.
We discuss the possibility that the universe is a neural network. The two types of degrees of freedom are bias and hidden.
Vanchurin tries to explain away the gap between classical and quantum physics. We know that quantum physics is able to explain what is happening in the universe at very small scales. We can use quantum mechanics when we deal with individual photons.
We lose the thread when we transition from observable quantum phenomena to classical observations, so we have to use classical physics to describe what is happening.
The argument is about something.
The root problem with trying to figure out a theory of everything is that it usually ends up replacing one proxy-for-god with another. The two most enduring explanations for our universe are based on different interpretations of quantum mechanics. Vanchurin attempts to reconcile with his "world as a neural network" theory by using the "many worlds" and "hidden variables" interpretations.
Vanchurin concludes:
In this paper, we discussed the possibility that the universe is a neural network. This is a very bold claim. Artificial neural networks can be useful for analyzing physical systems or for discovering physical laws, but we are saying that this is how the world around us actually works. It should be easy to prove it wrong, as it could be considered a proposal for the theory of everything. The only thing needed is a physical phenomenon that cannot be described by neural networks. It's easier said than done.
Vanchurin specifically says he is not adding anything to the "many worlds" interpretation, but that is where the most interesting philosophical implications lie.
If Vanchurin's work pans out in peer review, we will have a found thread to pull on that could put us on the path to a successful theory.
What is the network's purpose if we are all in it? Is the universe a closed network or a grander network? Maybe we are one of trillions of other universes connected to the same network. When we train our neural networks, we run a lot of cycles. Are we just one of many training cycles for a bigger machine?
You can read the whole paper on arXiv.
