Dr. Krenn’s research focuses on the development of digital discovery frameworks capable of designing novel physical experiments. In doing so, he questions whether human intuition may limit the discovery of innovative experimental methods. Through the targeted use of AI, virtual tools are used to recombine experimental components and find solutions not immediately understandable or intuitive to humans.

One example of this work is PyTheus, an open-source software for the digital discovery of new quantum experiments. PyTheus uses abstract mathematical graphs to solve complex quantum problems and has already discovered over 100 novel quantum experiments. “This is definitely a component of the future of science. Humans may not discover everything, much will be discovered by machines and humans will try to understand what the machine did,” says Krenn.

In the interview, Krenn talks about the challenges and successes of his research. One of the most significant discoveries made by PyTheus relates to the problem of quantum entanglement generation. Normally, entangled particles are generated at one location. Entanglement swapping is a method in which two previously unentangled particles become entangled through an indirect connection, without directly interacting. This technique is particularly important for quantum communication and quantum networks. While conventional methods for creating quantum entanglement rely on pre-existing entanglement, PyTheus has found alternative ways which do not require initial entanglement. These new approaches could fundamentally change our understanding of quantum entanglement.

In addition, Krenn’s group is working with the LIGO collaboration – the team which provided the first direct detection of gravitational waves. In this context, his team is exploring how AI can be used to develop novel detector designs which go beyond human intuition. Some of the AI-generated designs significantly outperform traditional human designs.

“One of the most interesting things I can think of in the future is building algorithms that can basically simulate all of experimental physics. So your AI system not only has access to all experimental equipment in a virtual way, but also to a huge number of open experimental physics questions. It’s likely to find some very unorthodox solutions,” says Krenn and adds: “I think about this as one of the most exciting large-scale collaborations between the domain of physics and the domain of artificial intelligence. That would be a dream to have that.”

About Quanta Magazine

›Quanta Magazine‹ is a leading science publication specializing in making complex scientific topics understandable to a broad audience. Known for its in-depth analysis and high-quality content, the magazine provides a platform for scientists to present their groundbreaking research to a global audience.

The full interview with Dr. Mario Krenn offers fascinating insights into his research and can be viewed here.