OpenAI has published findings on how more than 1,800 scientists use ChatGPT in their research. The report describes the ways the tool is integrated into daily research workflows, the productivity effects that scientists report, and their expectations regarding scientific discovery.
Why it matters
Understanding how a large group of scientists apply a general-purpose language model to research tasks offers insight into the practical role such tools may play in scientific work.
Who should care
Researchers, scientists, and others interested in how AI tools are being incorporated into research practice may find the reported workflows, productivity observations, and discovery expectations relevant.