Did This AI Just SOLVE Quantum Physics? | Unveiled

<h4>

Did This AI Just Solve Quantum Physics?</h4>

Artificial intelligence and quantum physics. They are two of the newest frontiers in science, promising to change the world forever. But, have we already discovered a way to understand one through the other?

This is Unveiled, and today we’re answering the extraordinary question; did this AI just solve quantum physics?

In all aspects of life, we’re always looking for a simpler solution. A way to do things that’s more efficient and less taxing, at the same time. And, at the cutting edge of the quantum realm, pioneering researchers are really no different. It’s only within the last few decades that we’ve even begun to understand subatomic matter… but, now, we want to properly control it. The problem is that that’s extremely difficult to do. After a fresh study turned in earnest to machine learning, though, many of the inherent problems may well have been resolved.

Published in “Physical Review Letters” in September 2022, the study in question is titled “Deep Learning the Functional Renormalization Group”. If that sounds quite complex, that’s because it is… there’s some serious work behind it, drilling down into the very fundamentals of life, the universe, and everything.

Led by Professor Domenico Di Sante of the University of Bologna and in collaboration with the Flatiron Institute’s Center for Computational Quantum Physics, in New York, the study started with a massive quantum problem requiring upwards of 100,000 separate equations to solve. When machine learning was applied to that problem, though, it reduced it down to just four equations, only. It turned a once impossible task into something that could be easily completed without fuss. It made what was once inconceivable into something that could be quickly and accurately understood.

The problem itself was played out on what’s known as the Hubbard Model. This involves monitoring the flow of electrons over a uniformed, grid-like lattice. It’s described as the idealization of how many quantum materials work, providing scientists with an optimum testing ground to view and record how and where electrons move. This is important because it’s knowledge that can then be applied to the construction of things like superconductors; to the creation of effectively custom-built materials, that can then be used for things like energy distribution. But, also, the strategy might be suitable for improving our understanding of our own bodies, too, of our brains, and to better connect us to the universe on a cosmological level.

But, back to the here and now, and so far we’re not quite at the stage of actually reaching a higher dimension. Nevertheless, the results of the study are game changing. The Hubbard Model lattice provides countless sites at which electrons can (and do) interact. We know, however, that through quantum processes including entanglement and superposition, the ways in which electrons can interact are very far from black and white. There are endless possible variations to consider, which is why even this relatively simple and idealized quantum arrangement will ordinarily leave us with hundreds of thousands of equations to sift through - if we ever hope to make sense of it. 

A Renormalization Group (as per the study’s complex title) is the means by which scientists compare and contrast activity in a quantum system, based on different variables. It’s what actually produces all those hundreds of thousands of equations that we’re wrestling with. Now, though, the deep (or machine) learning approach reduces all of that down to just four, granting researchers almost total control, knowledge and power over the system (or material) that they’re studying. It’s like we’ve tapped into the quantum level, and have suddenly translated it into something we really can understand - all thanks to some AI middle management. 

It’s still early technology, but it could turn out to have an almost Rosetta Stone-like impact, opening the door for more and more study. For an unassisted human being to achieve the same thing - to convert endless problems into a digestible few - it could take years, or even multiple lifetimes. But, for a powerful enough computer, it’s infinitely quicker. This primary study did reportedly take a few weeks to complete… but those behind it claim that all subsequent studies should be much more rapid, now that the base program is built and working. In the future, then, perhaps we’ll see near instant results, allowing us to wield quantum control in real time.

So, what does a future like that look like? In some ways, it could be pushing us closer to real life magic or superpowers than we’ve ever been before. Any society with quantum control is able to view all the matter that exists around (and inside) itself as malleable, non-final and therefore reformable. Instead of trialing and testing to find the best materials available for any one job, it can specifically build materials that do exactly what’s required. With full quantum control, then - which importantly isn’t yet what this experiment has promised - the applications are arguably limitless. If you had the power, what types of materials and products would you conjure first of all? How would you choose to alter the world, if you could dabble in the quantum domain?

There is still a way to go before we reach these higher stages. This recent study is more a starting point on the path toward that end goal, rather than the end goal itself. And, indeed, there are many other directions that it could take. But still, in a world that’s constantly demanding for more of everything - more speed, reliability, innovation, etc. - it’s an early breakthrough that can capture imagination. Again, we’re talking here about hundreds of thousands of previous problems… shrunken down to just four. At the very least, tracking the quantum world has never been more straightforward. Quite what it will mean for the future remains to be seen.

We could view this story in another way, too; as a major milestone reached as we move along the scale of microdimensional mastery. This was laid out by the British physicist and theoretical thinker, John D. Barrow, and is otherwise known as the Barrow Scale. It works a little like a reverse Kardashev Scale, where the emphasis is on gaining dominion over increasingly small things (like atoms and nuclei) as opposed to increasingly large things (like Kardashev’s star systems and galaxies). 

At our current level, humans exist somewhere between type one minus and type two minus on the Barrow Scale; we have a good grasp of the subatomic world and of our own genes, and we’re beginning to manipulate both, but our attempts so far are limited. However, by adding AI into the mix, and thereby reducing quantum problems so dramatically, we could soon be propelling ourselves onto type three minus, and beyond. This means having total power to control and distort at the molecular and subatomic levels; again to create new materials, but also as a precursor toward true nanotechnology. Machines so small that we can’t physically see them, and yet they could potentially be so vital to our continued evolution. At the very top tier of microdimensional mastery, we’re talking full control over time and space itself… but perhaps let's not get too far ahead of ourselves! Atomic level tinkering before total time travel… that’s the natural order of things, right?

What’s your verdict on the blending of AI with quantum physics? Are you hopeful for the future, unconvinced by some of the predictions, or maybe even fearful of the direction down which we’re seemingly headed? Both fields have certainly stoked debate in recent times. But, while the mainstream focus has been more on the advent of quantum computing outright, here we see how artificial intelligence really could have a hand in our physical reality - and maybe even very soon. 

For now, this is still just one experiment, conducted across an idealized scenario, involving a small sample of carefully shepherded electrons. It isn’t yet the answer to life, the universe and everything. But it does encourage us all to look at (and think about) our surroundings in a much deeper way. There are still many mysteries afoot, so we can’t truly say that this has solved quantum physics… but it has presented it in an all new and infinitely more accessible way. The AI approach could well prove crucial as we continue to progress.