Artificial intelligence is part of our modern world. The key question for practical applications is how fast such intelligent machines can learn. The experiment answered this question, showing that quantum technology makes it possible to speed up the learning process. Physicists have achieved this result by using a single-photon quantum processor as a robot.
Resolving computer games, recognizing human voices, or helping to find optimal medical treatments: these are just a few amazing examples of what the field of artificial intelligence has produced in recent years. The ongoing race for better machines has led to the question of how and with what means improvements can be made. At the same time, huge recent advances in quantum technologies have confirmed the power of quantum physics, not only for its often odd and puzzling theories, but also for real-life applications.
Hence, the idea of merging the two fields: on one hand, artificial intelligence with its autonomous machines; on the other hand, quantum physics with its powerful algorithms.
Intelligent machines can learn faster with quantum technology. An experiment has answered this question, showing that quantum technology enables a speed-up in the learning process.
Over the last few years, many scientists have begun to explore how to bridge these two worlds, and to study how quantum mechanics can prove beneficial to robot learning, or vice versa. Several fascinating results have been shown, for example, by robots deciding faster on their next move, or by designing new quantum experiments using specific learning techniques. But robots were still unable to learn faster, a key element in the development of increasingly complex autonomous machines.
Within the framework of an international collaboration led by Philip Walther, a team of experimental physicists from the University of Vienna, together with theoreticians from the University of Innsbruck, the Austrian Academy of Sciences, the University of Leiden and the German Aerospace Center, have been able to prove for the first time experimentally the speed of robot learning. The team used single photons, the fundamental particles of light, coupled to an integrated photonic quantum processor designed by the Massachusetts Institute of Technology.
This processor has been used as a robot and has been used to perform learning tasks. Here, the robot would learn how to route single photons in a predefined direction. “The experiment could show that learning time is significantly reduced compared to the case where quantum physics is not used,” says Valeria Saggio, the first author of the publication.
In a nutshell, the experiment can be understood by imagining a robot standing at a crossroads, with the task of learning to always turn left. The robot learns to make the right move by obtaining a reward. Now, if the robot is placed in our usual classical world, it will try either a left or a right turn, and it will be rewarded only if the left turn is chosen. By contrast, when the robot exploits quantum technology, the bizarre aspects of quantum physics come into play. The robot can now make use of one of its most famous and peculiar features, the so-called superposition principle.
This can be understood intuitively by imagining a robot taking two turns, left and right, at the same time. “This key feature enables the implementation of a quantum search algorithm that reduces the number of tests needed to learn the correct path. As a result, an agent that can explore its environment in superposition will learn significantly faster than its classical counterpart “According to Hans Briegel, who worked with his group at the University of Innsbruck to develop theoretical ideas on quantum learning agents.
This experimental demonstration that machine learning can be enhanced by using quantum computing has promising advantages when combining these two technologies. “We are just beginning to understand the potential of quantum artificial intelligence,” says Philip Walther, “and thus every new experimental result contributes to the development of this field, which is currently seen as one of the most fertile areas for quantum computing.”