Quantum mechanics isn’t easy. Sometimes, even scientists find it frustrating.
Quantum physicist Mario Krenn and his colleagues in the group of Anton Zeilinger, from the Faculty of Physics at the University of Vienna and the Austrian Academy of Sciences, had the counterintuitive features of quantum mechanics in mind when they developed “Melvin.”
Melvin is an algorithm which designs new useful quantum experiments. As the computer does not rely on human intuition, it finds novel and unfamiliar solutions.
The idea for Melvin came about when physicists were trying to create new quantum states in the laboratory, but just couldn’t seem to come up with viable methods to do it.
Typically, experiments tend to use a rather limited set of components, such as beam splitters, mirrors, and crystals. Melvin is able to take these elements and shuffle them around to find a new arrangement.
“After many unsuccessful attempts to come up with an experimental implementation, we came to the conclusion that our intuition about these phenomena seems to be wrong. We realized that in the end we were just trying random arrangements of quantum building blocks. And that is what a computer can do as well – but thousands of times faster,” explains Mario Krenn, lead author of the new research.
After just a few hours of calculation, the unfortunately named Melvin, a sophisticated heuristic algorithm, found the correct solution to the question that had stumped the researchers—and the result surprised them.
“Suppose I want to build an experiment realizing a specific quantum state I am interested in,” Zeilinger says. “Then humans intuitively consider setups reflecting the symmetries of the state. Yet Melvin found out that the most simple realization can be asymmetric and therefore counterintuitive. A human would probably never come up with that solution.”
In other words, this embryonic machine intelligence represents the ultimate in thinking “outside the box”—its ideas and solutions are unbounded by human mental and perceptual limitations. Which, very naturally, is a great asset in understanding the very non-intuitive world of the quantum domain.
After arranging the elements randomly, Melvin calculates the resulting photon beams and checks whether any of them comes close to the specified goals. If not, the process is repeated. But if the output meets the design criteria, Melvin then does some further shuffling to simplify the setup before reporting it to the user.
Quantum theory verifies that Melvin’s arrangements should work, but that doesn’t mean it’s easy to understand why (again, even for scientists).
“The solutions are difficult to understand, but we were able to extract some new experimental tricks we have not thought of before. Some of these computer-designed experiments are being built at the moment in our laboratories,” says Krenn.
Luckily for scientists, Melvin is continuously learning. He’ll continue to baffle them as he becomes smarter with each successful attempt. This significantly speeds up the discovery rate for more complex solutions.
In the future, the authors want to apply their algorithm to even more general questions in quantum physics, and hope it helps to investigate new phenomena in laboratories. However, they don’t see it replacing the human role in these experiments, simply augmenting it.
Hopefully, the program will limit its ambitions to designing quantum experiments, and will never surpass human intelligence—because if not, we’ll all rue the day when we have to bend the knee to our new computer overlord named…Melvin.