Is AI Advanced Enough To Do Science?

The last few decades have been great for artificial intelligence and machine-learning technology, advancing leaps and bounds and revolutionizing various sectors of society, from health care, entertainment to being integrated on your pocket smartphone. Some of these even, have been able to surpass their human counterparts, which is one of the reasons as to why automation is fast becoming a thing.

But can they do science?

That’s what a team of scientists based at the Okinawa Institute of Science and Technology Graduate University (OIST), the University of Munich and the CNRS at the University of Bordeaux discovered, when AI was able to solve complex problems just as accurately as scientists, essentially beating  theoretical physicists at their own game. In fact, the machines were even able to do it much faster.

Recently published in Physical Review B, the study showed a machine that was able to identify unusual magnetic phases in a model of pyrochlore -- a naturally-occurring mineral with a tetrahedral lattice structure. What’s remarkable, however, is while the OIST scientists needed six years to solve the problem, the machine only needed a few weeks.

"This feels like a really significant step. Computers are now able to carry out science in a very meaningful way and tackle problems that have long frustrated scientists," Professor Nic Shannon, who leads the Theory of Quantum Matter (TQM) Unit at OIST, said.

The problem in question is related to magnets and how piecing together the phase diagram and identifying the rules governing the interactions between magnetic substances in different phases can be hard. As such, the scientists were curious whether the TQM unit can solve it.

"To be honest, I was fairly sure that the machine would fail. This is the first time I've been shocked by a result -- I've been surprised, I've been happy, but never shocked," Prof. Shannon added. 

"We are thrilled by the success of the machine, which could have huge implications for theoretical physics. The next step will be to give the machine an even more difficult problem, that humans haven't managed to solve yet, and see whether the machine can do better."

Artificial Intelligence Scientists continue to explore artificial intelligence and how it could improve common processes, such as diagnosing diseases. Pixabay