The Importance of Open Source Science in the AI Era

ORAL  · Invited

Abstract

The field of nuclear physics stands at the precipice of a transformative era driven by artificial intelligence (AI) and machine learning (ML). From large-scale data analysis to complex simulations, AI/ML offers the potential to revolutionize our understanding of the fundamental building blocks of matter. However, to fully harness this potential, a paradigm shift towards accessibility and open science practices is essential. We will discuss the challenges and opportunities associated with sharing data, code, and workflows openly. Specific examples will be drawn from ongoing research efforts to illustrate how open source principles can accelerate scientific discovery and collaboration in the nuclear physics domain. Furthermore, the talk will address the unique considerations for open science in the context of AI/ML. The inherent complexities of AI/ML models and the potential for bias necessitate careful attention to transparency and reproducibility. We will explore strategies to ensure responsible and trustworthy use of AI/ML within an open science framework. By fostering a culture of openness and collaboration, the nuclear physics community can leverage the power of novel computational paradigms to unlock new frontiers in our understanding of the nucleus while making it more accessible than ever to get involved in forefront nuclear science.

*This work is supported in part by funding from the National Science Foundation under award number 2004601 (CSSI program, BAND collaboration) and the U.S. Department of Energy under Awards Nos. DE-SC0023688 and DE-SC0023175 (NUCLEI SciDAC-5 collaboration).

Presenters

  • Kyle S Godbey

    • Michigan State University
    • FRIB, Michigan State University
    • Facility for Rare Isotope Beams

Authors

  • Kyle S Godbey

    • Michigan State University
    • FRIB, Michigan State University
    • Facility for Rare Isotope Beams