Tech

Physicists working with Microsoft compare the universe to an AI

The laws of physics are the result of a long learning process on the part of the universe. Here is the theory presented by several scientists, which thus compares our world to a computer endowed with machine learning. A study that turns the environment upside down since it amounts to saying that we will never unlock all the secrets of our environment.

universe
Credits: Pixabay

What if we all lived in a simulation? If The Matrix did not invent this theory, the film of the Wachowski sisters undoubtedly popularized it. One might think that only a few enlightened individuals believe in it, and that the scientific community would only look at it amused. However, there is nothing far-fetched about compare our universe to a computer. It is even on this principle that a recent study published by a team of physicists is based.

According to them, all the laws of physics are the result oflearning by the universe. In other words, we would be living in some sort of huge computer system operating like an AI. It may sound anecdotal, but it is not. This would mean that the laws that govern our world are not fixed and, on the contrary, they evolve over time. In other words, we will never unravel the mysteries of the universe.

On the same subject: This artificial intelligence allows you to chat with a virtual Einstein

The universe as an AI, a theory that revolutionizes physics

Indeed, as Stephen Hawking writes in his cult “A brief history of time”, the ultimate quest of physicists has always been to theorize THE rule that explains our universe. However, if the latter is constantly learning and evolving, what is true today may not have been true yesterday, and may not be tomorrow. “For example, when we see structures that resemble deep learning architectures emerge in simple self-study systems, could we imagine that the operational matrix architecture in which our universe creates laws, has itself evolved from it? ” a self-study system[…]? “, thus asks the study.

In this way, it would mean that our universe first experienced an initial state with “The most minimal possible starting conditions”. Evolving then like a deep learning algorithm, the latter would have perfected its laws, or even created new ones. according to the results obtained by the previous ones. As we describe these laws as we observe them at a time T, it seems logical that the very first of them was particularly “simplistic”. In addition to shake up the world of physics, this theory has other important consequences.

On the same subject: Artificial intelligence will soon be able to evolve like a living being

The universe might never turn back

“One of the consequences is that if the evolution of laws is real, it is likely that it is unidirectional, because otherwise it would be common for laws to revert to previous states, perhaps even more likely to find a new state “, continue the physicists. “This is explained by the fact that a new state is not random but must rather meet certain constraints, whereas the immediate past state has already met the constraints. A reversible but evolving system would frequently explore its immediate past in a random fashion. “

Here again, the study goes against other theories, who predict that the universe will eventually return to a density level close to that of the Big Bang. To explain their point of view, scientists use the example of software. By examining the hard drive, it is possible to determine the actions performed by the program by consulting the magnetic marks left: the results are therefore reversible. Conversely, using the processor, the same task will be impossible to perform, because no history will be kept.

It is therefore difficult to predict the impact that this study will have in the world of physics. But, if it turns out that the laws of relativity and even of gravity are not actually written in stone, it is a safe bet that such a theory will profoundly change our way of seeing the world.

Source

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *