Carnegie Mellon University
September 08, 2023

Physicist Looks to AI for New Answers to Universal Questions

By Heidi Opdyke

The magical mystery of quantum mechanics thrills Beka Modrekiladze. The Carnegie Mellon University physics doctoral student at Mellon College of Science sees artificial intelligence as the way to unlock its secrets.

“I want to understand what space-time is made of,” he said. “Is it fundamental, or is it something else? One of the most important goals in my life is to understand the universe on a deeper level.”

Quantum gravity is one way that Modrekiladze studies space-time, a model that joins the two concepts into a single idea.

“Space-time oscillates, and we can measure this oscillation to see how time slows down or accelerates, and we can detect those waveforms,” he said. One challenge is that there is a lot of noise, so researchers rely on looking for waves similar to pre-existing candidates.

“That approach was philosophically unsatisfying for me,” Modrekiladze said. “To really accelerate science, we have to think beyond our imaginations.”

That’s where AI comes in.

“The universe doesn’t always make sense,” Modrekiladze said. “But mathematics helps us to close our eyes and follow its logic to get somewhere new.”

“Mathematics sometimes deletes bias. That’s how quantum mechanics was discovered. We followed the mathematics to see that the universe works in an unexpected way. AI is hyper-mathematics in that sense, because it will push us out of our comfort zones to make even more new discoveries.”

Carnegie Mellon is considered the birthplace of AI. As a member of Carnegie Mellon’s faculty for more than 50 years, Nobel Prize Laureate Herbert “Herb” A. Simon conducted extensive research on human decision-making and problem-solving processes, which led to the creation of computational tools that both simulated and augmented human thinking. Simon’s work with Allen Newell led to the invention of computer systems that could learn and adapt, creating the modern-day field of artificial intelligence.

“Carnegie Mellon was the only choice for my Ph.D., because it has a really good physics department, and it is literally the best at AI in the world,” Modrekiladze said.

“Mathematics sometimes deletes bias. That’s how quantum mechanics was discovered. We followed the mathematics to see that the universe works in an unexpected way. AI is hyper-mathematics in that sense, because it will push us out of our comfort zones to make even more new discoveries.”

Beka Modrekiladze
physics doctoral student, Mellon college of science

Modrekiladze published a paper in the Journal of High Energy Physics earlier this year on gravitational wave signals from the collision of black holes and compact stars.

“It’s the ultimate laboratory that the universe can offer,” Modrekiladze said about the phenomenon. “The high energies are the perfect place to test our best theories. Being able to study gravitational waves, which are the oscillations of space and time itself, introduced a new era for astronomy.”

Through the paper, Modrekiladze expanded the framework that can predict what type of gravitational waveforms researchers can expect when two black holes or a black hole and other exotic star material collide. As one of the applications of the paper, a new type of gravitational signals, potentially opening new physics, will not be missed.

“Currently, we can detect gravitational wave signals, if their waveforms are already in the catalog,” he said. “To go beyond our expectations, the next step for me was to create generative AI, which would generate new types of possible waveforms.”

Through studying gravitation waves, researchers can use mathematical models to see different properties and possibly distinguish black holes from exotic stars.

“It could answer the question: Does space-time preserve parity?” he said. “A simpler way to say this is, if we show the universe itself in a mirror, can it tell the difference? We don’t know if space-time has that symmetry or not.”

“There are great sci-fi books that use ideas of general relativity. But there are not many books using the magic and craziness of quantum mechanics. It’s usually hard to convey in a story, but at the same time, I had philosophical ideas I wanted to communicate and sci-fi was a framework where I could do that.”

Beka Modrekiladze
physics doctoral student, Mellon college of science

To help him think through his own philosophy and ideas around AI and quantum mechanics, Modrekiladze uses storytelling. In 2019, he published “Eternal Gods Die Too Soon,” in which the main character is an AI entity similar to ChatGPT that wasn’t released until November 2022.

“There are great sci-fi books that use ideas of general relativity,” he said. “But there are not many books using the magic and craziness of quantum mechanics. It’s usually hard to convey in a story, but at the same time, I had philosophical ideas I wanted to communicate and sci-fi was a framework where I could do that.”

“Probably the most important part in that book for me was trying to answer the question, is mathematics discovered or invented? On the one hand, ancient Greeks wrote formulas, and it was our creation. But then we discovered that the entire universe follows our invention. And not only now, but before people were created. That raised really big questions.”

Physics Professor Ira Rothstein advises Modrekiladze and said the broad interests his student brings in physics, computer science and other disciplines offers a lively perspective for their work.

“He’s just a pleasure to have around,” Rothstein said. “He makes the atmosphere exciting and energetic.”

A native of Tbilisi, Georgia, Modrekiladze was recognized by his home country as a Tsinandali Award distinguished nominee in 2022. The award honors Georgia’s brightest young talents in arts and sciences.

“It’s a prestigious award in Georgia,” he said. “It’s a reassurance that what I’m doing is important.”