Artificial Intelligence As a Driver of Modern Science: From Biomedicine to International Megascience Projects
摘要
This review summarizes the key developments of the past two decades, highlighting the transformation of scientific research driven by artificial intelligence (AI). It explores both well-established and emerging applications of AI—ranging from biomedicine, pharmaceuticals, nutrition science, education, data science, engineering, and fundamental disciplines to synchrotron radiation facilities and advanced megascience infrastructures. The paper provides a detailed overview of AI approaches and architectures implemented in major international projects, including the Large Hadron Collider and its High-Luminosity upgrade, the FAIR facility, the DUNE and Belle II experiments, the LIGO, Virgo, and KAGRA gravitational-wave observatories, the LSST survey telescope, the SKA radio interferometer, the Euclid space observatory, the ESS, SNS, and J-PARC neutron sources, the ITER and JT-60SA fusion reactors, the NIF laser facility, the ELI-NP laser complex, exascale climate simulation platforms, and cryo-electron microscopy pipelines powered by AlphaFold technology. The article is intended for a broad audience of specialists interested in the methodological, technical, and organizational aspects of AI integration into scientific research.