Can a computer algorithm discover something new about physics? It is a fascinating question. a new research work on the subject inspired the sensational headline “An AI may have invented ‘alternative’ physics.”
The term “alternative physics” sounds a lot like “alternative facts”, but let’s investigate anyway. How does the performance of this computer program compare to that of a real physicist? Or even that of an average student?
Isaac Newton was a peerless genius. The English scholar not only unified the studies of motion and gravity, but also invented the mathematical language with which to describe them. The concepts of classical mechanics created by Newton are the basis of most of the physics invented since then. His concepts were later reformulated in a new mathematical language in the eighteenth century by the exceptional continental physicists Joseph-Louis Lagrange and Leonhard Euler.
Newtonian mechanics requires an analysis of directional forces acting on massive bodies. If you took an introductory physics class in high school or college, you’ve seen these problems: boxes on inclines, pulleys, and carts. You draw arrows going in various directions and try to balance the forces. It works great for small problems. As problems get more complex, this method continues to work, but it gets brutally tedious.
With Lagrange’s formulation, if two aspects of the nature of the system can be defined, the problem can be solved using calculus alone. (Yes, “just” calculus: Processing derivatives is much easier than solving extremely complex free-body diagrams where the arrows change at every position.)
The first thing to understand is the energy of the system, that is, the (kinetic) energy of motion and the (potential) energy stored by the configuration of the system. The second crucial thing is to choose the appropriate coordinates, or variables, for the motion of the system.
Subscribe to get amazing, surprising and shocking stories delivered to your inbox every Thursday
Imagine a simple pendulum, like that of an old clock. The pendulum bob has kinetic energy due to its swinging motion and potential energy due to its position (height) within the gravitational field. The position of the pendulum can be described by a single variable: its angle relative to the vertical. The Lagrange solution for the motion of the pendulum can be calculated with relative ease.
Solving more complex problems in mechanics requires discovering the right number of variables that can describe the system. In simple cases this is easy; in moderately complex cases, it is a student-level exercise. In extremely complex systems, it can be the work of a professional or impossible. This is where the “physical” AI comes in.
AI physicist is defeated by college students
The computer was configured to analyze the problem of a pendulum hanging from another pendulum. This problem requires two variables, the angle of each pendulum with the vertical, or four variables if a Cartesian (xy) coordinate system is used. If both weights of the pendulum are hanging on springs instead of rigid rods, the two variable spring lengths are added together to get six variables in the Cartesian system.
The computer was asked to determine the number of variables needed to compute the above problems. How did the AI physique do? Not good. For the rigid pendulum on a pendulum he gave two answers: ~7 and ~4-5. (The correct answer is 4 variables). Similarly, he calculated ~8 and ~5-6 for the double spring pendulum. (The correct answer is 6 variables). The researchers praise the smaller estimates for being close to the true answers.
But after digging into the details of the article supplementary materialsHowever, the result begins to unravel. The computer didn’t actually calculate 4 variables and 6 variables. His best estimates were 4.71 and 5.34. None of those answers even round to the correct answer. The four-variable problem is an intermediate undergraduate physics problem, while the six-variable problem is a more advanced undergraduate problem. In other words, the average physics college student is significantly better than the AI physicist at understanding these problems.
AI physique not ready for tenure
The researchers continue to ask the program to analyze complicated systems that not only have an unknown number of variables, but it is not clear whether classical mechanics can describe the systems at all. Examples include a lava lamp and fire. The AI does a decent job of predicting small changes to these systems. It also calculates the number of variables required (7.89 and 24.70, respectively). The correct answers to these problems would be “new physics”, in a sense, but there is no way of knowing if the AI is correct.
Using AI to analyze unknown systems is a good idea, but AI currently can’t get the easy answers right. Therefore, we have no reason to believe that he is hitting the hard ones.
#discover #type #physics