AI is a part of the physical world just like us, but it looks like it has something personal to do with physics. A few months ago, AI made its first step towards altering the way we see physics. Yes, you heard it right—a new AI program developed by researchers at Columbia University discovered its own alternative physics. After being shown videos of physical phenomena on Earth, the AI didn’t rediscover the current variables we use; instead, it came up with new variables to explain what it saw.
In recent times, researchers from Flatiron’s Center for Computational Quantum Physics (CCQ) have shown that by using neural networks to reduce the mathematical representation used to describe a quantum system, they may learn a great deal more about it. Instead of just presenting physics and forecasting results for other scientists to find, the new technique makes it easier to find hidden patterns. Keep reading to find out more about this amazing achievement.
Sir Isaac Newton’s groundbreaking work in physics was first published in 1687 in his book “The Mathematical Principles of Natural Philosophy,” commonly known as “The Principia.” In it, he outlined theories about gravity and motion. Hooke’s Law, for example, states that the extension of a spring is proportional to the tension stretching it. A doubling of the tension results in a doubling of the amount of stretch. Or the law of conservation of energy: energy can neither be created nor destroyed.
AI to alter the way we perceive physics?
It wouldn’t be surprising. Here’s what the AI program developed by researchers at Columbia University did:
“In the experiments, the number of variables was the same each time the AI restarted. But the specific variables were different each time. ” “So yes, there are alternative ways to describe the universe, so it’s quite possible that our choices aren’t perfect.” Roboticist Hod Lipson from the Creative Machines Lab at Columbia
Certainly, there are multiple ways to perceive the physical world. Now we have AI to support the idea.
If extraterrestrials were to perceive physics, imagine how many different variables they would use to explain and characterize any phenomenon. Imagine the types of variables they could discover next, their parameters, and the laws they might arrive at.
We can only have one location that can describe the variables on which it operates, but not all of them. Is it possible that different universes each contain an infinite number of variables? Our physics is limited by our mind’s inability to describe nature beyond its reach in mathematical terms.
Will the physical laws apply to artificial intelligence?
The simple answer is yes; however, there are multiple questions. The question to which the answer was “yes” was whether or not the laws of physics would apply to artificial intelligence. A robot could defeat gravity while staying well within the laws and bounds of nature.
For example, the “soul” of a physical robot powered by AI could reside inside an anti-gravity machine. As such, the robot could even walk on water. The robot would possess both physical and nonphysical properties. This can be further explored by using what we call “counterfactual reasoning. That’s “If a statement is true, then under certain conditions, it is false.”
In this case, we’d be trying to assess whether or not the robot that had part of its existence in an anti-gravity field could also have its other half in another dimension. This means that the anti-gravity body could’ve been created in a laboratory by humans, while the non-physical part of it exists in another world altogether.
AI could help us explore other dimensions
We have been living, or as some say, “trapped”, in a three-dimensional reality with height, width, and depth. Less obviously, since we can only sense one aspect of it, we can consider time to be the fourth dimension, as Einstein famously revealed.
As was already said, its length is defined by its first dimension (aka. the x-axis). A straight line is a good example of a one-dimensional object. It’s because it has only one dimension, length, and no other identifiable features.
The object becomes a 2-dimensional shape when the y-axis (or height) is introduced as a second dimension. The third dimension involves depth (the z-axis) and gives all objects a sense of area and a cross-section. And only one aspect of the time dimension.
Philosophers have proposed the possibility of many other dimensions than those we are living in.
The 26 dimensions of closed, unoriented bosonic string theory are understood as the 26 dimensions of the traceless Jordan algebra J3(O)o of 3×3 octonionic matrices, with each of the three octonionic dimensions of J3(O)o, whereas superstring theory involves the existence of nine spatial dimensions and one-time dimension for a total of ten dimensions.
Although they could imagine so many probabilities beyond their capabilities, humans are unlikely to alter the laws of physics and ever grasp the other worlds on their own. But AI could change the way we use physics in terms of thinking about the things that are outside our deterministic world. An advanced, next-level AI, therefore, could be useful in diving into the ocean of mysteries, holding humans on their backs.