PreprintCC BY-NC-ND 4.0arXiv
AI in Particle Physics Education: Research Problems and Foundational Skills
Abstract
Generative AI can produce perfect solutions to standard physics homework, weakening the connection between submitted work and knowledge a student can use independently. We report a redesign of the introductory nuclear and particle physics course at Ruhr University Bochum, in which AI was permitted and encouraged alongside traditional tools on unusually difficult, research-shaped assignments. We found that most students remained engaged with the assignments throughout the course and produced ambitious work, while the course did not reliably secure its foundational objective. We offer this mixed experience as a reference point for discussion.
Related
- Brilliance: toward autonomous computational research in particle accelerator physics
- PINNIES: An efficient physics-informed neural network framework for integral operator problems
- Transitioning from A-level to higher education physics courses: where are the knowledge gaps?
- Track finding and particle identification performance with particle physics foundation models
- How particle physics stops ultra-relativistic bubbles and shapes their gravitational waves
BibTeX
@misc{mikhasenko2026particle,
title = {{AI in Particle Physics Education: Research Problems and Foundational Skills}},
author = {M. Mikhasenko and M. Stahl and I. Segal and D. Parmar and A. Zimmer and A. Kazatsky},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.05719},
eprint = {2609.05719},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.05719},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - AI in Particle Physics Education: Research Problems and Foundational Skills
AU - M. Mikhasenko
AU - M. Stahl
AU - I. Segal
AU - D. Parmar
AU - A. Zimmer
AU - A. Kazatsky
PY - 2026
DA - 2026/09/04
PB - arXiv
DO - 10.48550/arxiv.2609.05719
UR - https://arxiv.org/abs/2609.05719
AB - Generative AI can produce perfect solutions to standard physics homework, weakening the connection between submitted work and knowledge a student can use independently. We report a redesign of the introductory nuclear and particle physics course at Ruhr University Bochum, in which AI was permitted and encouraged alongside traditional tools on unusually difficult, research-shaped assignments. We found that most students remained engaged with the assignments throughout the course and produced ambitious work, while the course did not reliably secure its foundational objective. We offer this mixed experience as a reference point for discussion.
ER -