Short summary
As AI spreads through scientific research, journals, institutions and classrooms are establishing rules for the use of large language models. However, a preliminary study published on arXiv on July 1 that has not yet undergone peer review argues that a laboratory’s approach should be shaped by the research values the team prioritizes.
Virginia University remarks
The study, prepared by space science researchers, identifies four types of laboratories that can make use of AI at different stages of research, rather than proposing a single policy suitable for all groups. West Virginia University astronomer Sarah Burke-Spolaor, a member of the team, says people are already using the tools, adding that they would not try to tell them, “Do not use AI.”
Telescope Science echoes
Because of applications such as code generation and writing assistance for people whose native language is not English, the focus is shifting from AI to group values. A radar chart and worksheet prepared with contributions from Michelle Ntampaka of the Space Telescope Science Institute in Baltimore help laboratories determine their priorities.