Artificial vs Human Intelligence: Physics/Chemistry Nobel Prize 2024

The 2024 Nobel Prizes in Physics (here) and Chemistry (here) were both awarded to applications of Artificial Intelligence AI developed by Google evidently surpassing Human Intelligence HI. 

The Prizes signify a crisis of modern physics and chemistry based on HI mathematical modeling, as delivering too little. Here AI steps in with a radically different approach of finding patterns from big experimental data sets by computation, without mathematical modeling. Compare with Physics is Dying.

The basic mathematical model of modern physics is Schrödinger's Equation SE describing properties of atoms and molecules built from protons and electrons, as the subject of modern physics and chemistry over the last 100 years. SE for microscopics without gravitation is complemented by Einstein's Equations EE for gravitation on macroscopic scales supposed to replace Newton's Equations NE.

SE describes evolution in time of electron distributions as a wave function with 3N spatial dimensions for a system with N electrons. Computational solution of SE requires work growing exponentially with N and so is beyond the capacity of any thinkable computer already for N=10. Also EE are effectively uncomputable, while NE are computable for very complex systems.  The fundamental models of modern physics in the form of SE and EE are thus basically uncomputable which shakes the theoretical foundation of physics based on mathematical modelling. No wonder that physics is dying.   

The result is that SE can only be solved by HI finding smart ways to drastically bring down the computational work, but after harvesting catches over 100 years including many Nobel Prizes, it now seems that HI with SE can no longer deliver something worth of a Prize, which is instead awarded AI without SE. In particular, simulation of protein folding by SE is impossible.

A shift from HI with math model to AI by computation as in the 2024 Nobel Prizes represents a major shift away from of the rationality of the Enlightenment and Scientific Revolution, with the chemistry prize to Google's AlphaFold2 for protein folding.   

Is there any hope of come back of HI in the form of HI with computation? 

Yes, there is a new form of SE which is computable for many electrons (work scaling with N) described as Real Quantum Mechanics with many posts under tags RealQM and Real Quantum Chemistry. It is possible that protein folding can be simulated by RealQM. 

Further, it may well be (here) that NE captures all of gravitation in a computable model with thus RealQM + NE emerging as a computable foundation of physics.  

So maybe HI with computation can still compete with AI by computation, while a combination must be optimal. 

PS On computational vs pure mathematics:

AI of today appears as a form of machine learning of neuronal networks trained on large data sets using computational tools of optimisation.  

AI can thus be viewed to be a form of computational mathematics as brain+computer, to be compared with pure mathematics performed by pen on paper using only brains. 

Does AI signify the end of pure mathematics carried by smart brains in a new era dominated by AI brute force computation?

Is it possible that pure mathematics can be assisted by AI for checking of proofs by brain or even as smart assistant proving theorems beyond reach by brain?

How far can HI + computer reach?