Generative AI tools in the nuclear engineering community: A survey-based evaluation of the current adoption and impacts
Generative AI tools (often abbreviated genAI or GAI), such as chatGPT, have arguably great potential to induce a paradigm shift in knowledge work.Herein, we present the results of a survey conducted within the nuclear engineering professional community to assess the current status and impact of genAI tools.Our study involved participants from various geographical locations, predominantly the USA and western Europe, representing a balanced distribution across academia, research institutions, and industry.We observed a consistent genAI adoption rate across career stages, except for students who exhibited a notably high adoption rate.We investigated the tasks facilitated by genAI and the reasons behind its non-utilization.We also analyzed the overall impact on workflow efficiency, estimating an overall 5% acceleration within the nuclear engineering community, albeit with considerable variability reported by individuals.Primary concerns raised included the accuracy and verifiability of genAI results, alongside specific apprehensions regarding export control and data security.Despite these concerns, respondents generally expressed satisfaction with using genAI tools.The survey provides a first quantification attempt at understanding the extent of genAI adoption, the specific tasks benefiting from its implementation, and the current impact on work efficiency.
2-s2.0-85202803484
École Polytechnique Fédérale de Lausanne
École Polytechnique Fédérale de Lausanne
College of Engineering
2024
9780894487972
1457
1465
REVIEWED
EPFL
Event name | Event acronym | Event place | Event date |
San Francisco, United States | 2024-04-21 - 2024-04-24 | ||