#### It is not every day that we come across a newspaper that tries to redefine reality. It is the case of a provocative pre-press uploaded to arXiv this summer.

A physics professor at the University of Minnesota named Duluth **Vitaly Vanchurin** it attempts to reformulate reality in a particularly illuminating way. His paper suggests that we are living within a massive neural network that governs everything around us. In other words, there is a "possibility that the entire universe at its most fundamental level is a neural network".

For years, physicists have been trying to reconcile quantum mechanics and general relativity. One postulates that time is universal and absolute, the other holds that time is relative, tied to the fabric of space-time.

## The paper

In his article, Vanchurin argues that artificial neural networks can "exhibit approximate behaviors" of both universal theories. Since quantum mechanics "is a remarkably successful paradigm for modeling physical phenomena on a wide range of scales," he writes, "it is widely believed that at the most fundamental level the entire universe is governed by the rules of quantum mechanics and also the severity it should somehow emerge from it. "

## Is the entire universe a neural network?

“We are not just saying that artificial neural networks can be useful for analyzing physical systems or for discovering physical laws. We are saying that this is how the world around us works ”, reads the discussion of the document.

In this respect it could be considered as a proposal for the theory of everything, and as such it should be easy to prove

Vitaly Vanchurin

The concept is so bold that most physicists and machine learning experts contacted declined to comment, showing skepticism about the article's conclusions.

## Vanchurin's clarifications

“My paper argues that the universe may fundamentally be a neural network. How would I explain the reasoning to someone who didn't know much about neural networks or physics?

**There are two ways to answer the question.**

**The first way** it is to start from a precise model of neural networks and then to study the behavior of the network within the limit of a large number of neurons.

What I have shown is that the equations of quantum mechanics describe quite well the behavior of the system near equilibrium and the equations of classical mechanics describe quite well how the system is further from equilibrium. Coincidence? It may be, but as far as we know quantum mechanics and classical mechanics describe exactly how the physical world works.

**The second way** is to start from physics. We know that quantum mechanics works quite well on a small scale and general relativity works quite well on a large scale, but so far we have not been able to reconcile the two theories into a unified framework. This is known as the quantum gravity problem. Clearly, we are missing something big, but to make matters worse we don't even know how to handle observers. This is known as a measurement problem in the context of quantum mechanics and a measurement problem in the context of cosmology.

**So it could be argued that there are not two, but three phenomena that need to be unified: **quantum mechanics, general relativity and observers. "

99% of physicists would say that quantum mechanics is the main one and everything else should somehow emerge from it, but no one knows exactly how to do it.

## Vanchurin's research

In this article the professor considers, as mentioned, another possibility. The possibility that a microscopic neural network is the fundamental structure and everything else comes as a consequence: quantum mechanics, general relativity and macroscopic observers. Things look pretty promising so far.

## How was this idea born?

“First I just wanted to better understand how deep learning works and so I wrote an article entitled 'Towards a theory of machine learning'. The initial idea was to apply the methods of statistical mechanics to study the behavior of neural networks, but it has been found that in certain limits the learning (or training) dynamics of neural networks is very similar to the quantum dynamics we see in physics. At that point I decided to explore the idea that the physical world is actually a neural network. The idea is definitely crazy, but what if it were crazy enough to be true? "

And would it pair (disprove it or integrate it?) With the theory of simulation.