idw – Informationsdienst Wissenschaft

Nachrichten, Termine, Experten

Grafik: idw-Logo
Grafik: idw-Logo

idw - Informationsdienst
Wissenschaft

Science Video Project
idw-Abo

idw-News App:

AppStore

Google Play Store



Instanz:
Teilen: 
06.09.2024 15:54

1.5 million euros in funding for Mario Krenn's research into artificial intelligence in physics

Maximilian Rinke Kommunikation und Marketing
Max-Planck-Institut für die Physik des Lichts

    Researcher Dr. Mario Krenn is developing artificial intelligence (AI) to complement the ingenuity and creativity of human researchers. The researcher at the Max Planck Institute for the Science of Light in Erlangen has now been awarded an ERC Grant worth 1.5 million euros with the aim of accelerating the efficiency and speed of scientific discoveries in the field of quantum physics through AI. The ERC Grant is awarded to young researchers to support them with pioneering ideas.

    With the surge of new possibilities, AI is increasingly being applied in scientific research, from weather forecasting to microbiology. One application in physics is the automated development and design of new physics experiments. Typically, experiments are designed by experienced and creative researchers. However, when these experiments become too complex or, as in the case of quantum physics, highly unintuitive, computer algorithms can assist in generating new experimental ideas. One of the main challenges with AI-based design of new experiments in physics is that these algorithms need to understand how experiments should function in nature. Virtual simulations of experimental setups are used for this purpose. To design a new, useful physics experiment, the simulator must be called upon millions of times. Therefore, it is crucial that these simulators are both extremely fast and highly reliable.

    With his project ARTDISQ (Artificial Scientific Discovery of Advanced Quantum Hardware with high-performance Simulators), Mario Krenn proposes to entirely rewrite and expand the simulators using modern, highly efficient methods. He utilizes the high-performance language JAX, developed by the tech-giant Google, originally intended for training the company's neural networks. Instead of neural networks, Krenn now writes physical simulation software in this language, which his AI algorithms can access. In another microscopy project, this has resulted in a 10,000-fold acceleration. The goal is to harness this potential and apply it to general physics simulators.

    Krenn intends to use these simulators to design new quantum-enhanced hardware. “I hope that with the help of AI, I can develop new, unorthodox technologies that will allow us to make novel and much more sensitive observations of the universe. Therefore, we will focus primarily on new telescopes and gravitational wave detectors,” says Dr. Krenn about his research. The computer's solutions and designs are often very different from what humans propose. These different solutions often inspire human scientists to come up with new tricks, ideas and concepts that would not have occurred to them otherwise.

    The European Research Council (ERC) is a European funding organization for outstanding frontier research. It was established by the EU in 2007 as part of the Horizon Europe program. The ERC offers four central funding programs. These include the Starting Grants, which are intended to make it easier for promising projects by young researchers to get off the ground. The total budget of the program for the period from 2021 to 2027 amounts to more than 16 billion euros. This year alone, 494 Starting Grants were awarded to scientists across Europe. Each grant amounts to 1.5 million euros over a period of five years. “This funding allows me to implement an idea that – if we are successful – could fundamentally change the development of physics experiments. This will make it possible to observe our world in new ways and develop innovative technologies”, says Krenn.

    The application of AI in physics has occupied Dr. Mario Krenn for around ten years. In 2014, during his PhD as an experimental physicist in Anton Zeilinger's group in Vienna, he developed the idea of using AI to design new experiments in quantum physics. After his dissertation, he moved to Toronto for 2.5 years to join one of the pioneering groups for AI-based material design. There, he wanted to learn everything that chemists already know about artificial design and the discovery of molecules and materials in order to transfer this knowledge to physics. Since September 2021, he has led his own independent research group at MPL, the Artificial Scientist Lab, where he develops new methods and applications.


    Wissenschaftliche Ansprechpartner:

    Dr. Mario Krenn
    mario.krenn@mpl.mpg.de
    Max Planck Institute for the Science of Light
    Research group Mario Krenn ›Artificial Scientist Lab‹


    Bilder

    Space of experimental configurations is enormously large. So far, it has mainly been explored by experienced human researchers. In Krenn's ARTDISQ project, he will develop highly efficient AI-based discovery methods for new quantum physical experiments.
    Space of experimental configurations is enormously large. So far, it has mainly been explored by exp ...

    © Mario Krenn

    Dr Mario Krenn, research group leader ›Artificial Scientist Lab‹
    Dr Mario Krenn, research group leader ›Artificial Scientist Lab‹

    © Stefan Spangenberg


    Merkmale dieser Pressemitteilung:
    Journalisten
    Mathematik, Physik / Astronomie
    überregional
    Personalia
    Englisch


     

    Hilfe

    Die Suche / Erweiterte Suche im idw-Archiv
    Verknüpfungen

    Sie können Suchbegriffe mit und, oder und / oder nicht verknüpfen, z. B. Philo nicht logie.

    Klammern

    Verknüpfungen können Sie mit Klammern voneinander trennen, z. B. (Philo nicht logie) oder (Psycho und logie).

    Wortgruppen

    Zusammenhängende Worte werden als Wortgruppe gesucht, wenn Sie sie in Anführungsstriche setzen, z. B. „Bundesrepublik Deutschland“.

    Auswahlkriterien

    Die Erweiterte Suche können Sie auch nutzen, ohne Suchbegriffe einzugeben. Sie orientiert sich dann an den Kriterien, die Sie ausgewählt haben (z. B. nach dem Land oder dem Sachgebiet).

    Haben Sie in einer Kategorie kein Kriterium ausgewählt, wird die gesamte Kategorie durchsucht (z.B. alle Sachgebiete oder alle Länder).