Our research on ai and science
We have been using machine learning tools for nearly a decade to conduct award-winning research on how social media platforms impact human behavior. Part of this work involved interfacing with Twitter and Meta, and we learned the hard way that the goals of industry and the goals of science are not always aligned. The communal values of science have always been in tension with private commercial interests, and we cannot expect private companies to provide reliable access to data or infrastructure for scientific research if it does not serve their interests.
In collaboration with experts in science and technology studies, we have led conversations within and beyond cognitive science about how AI products are changing the nature of scientific work. Leveraging the benefits of automation for scientific work does not require us to cede our agency to commercial products or industry interests. We have developed best practices in our lab for approaching automated workflows with thoughtfulness, built pipelines for analyzing large text corpora with open-source LLMs, and led workshops to educate our colleagues on these practices.
For more details on how we use automation in our research, please see our lab statement on AI and scientific research, a resource on the ethical costs and epistemic risks of LLMs, and our published research on AI and science:
Proprietary LLMs impede scientific transparency and reproducibility
McLoughlin, K.L., Palmer, A., Matias J.N., & Crockett, M.J. (2026)
Nature Reviews Psychology. Link» View PDF»
A reporting checklist for large language models in behavioural science
Feuerriegel, S., Barrie, C., Crockett, M.J., Globig, L.K., McLoughlin, K.L., et al. (2026)
Nature Human Behavior. Link» View PDF»
The uncritical adoption of AI in science is alarming — we urgently need guard rails
Messeri, L. & Crockett, M.J. (2026)
Nature. Link» View PDF»
Making Automation Work for Social Scientists
Crockett, M.J. (2026)
Dædalus. Link» View PDF»
Producing more while understanding less with large language models
Crockett, M. J. & Messeri, L. (2026)
Trends in Cognitive Sciences. Link» View PDF»
AI Surrogates and illusions of generalizability in cognitive science
Crockett, M.J. & Messeri, L. (2025)
Trends in Cognitive Sciences. Link» View PDF»
Artificial intelligence and illusions of understanding in scientific research
Messeri, L. & Crockett, M.J. (2024)
Nature. Link» View PDF»
Science communication with generative AI
Alvarez, A., Caliskan, A., Crockett, M.J., Ho, S.S., Messeri, L., & West, J. (2024)
Nature Human Behavior. Link» View PDF»
The limitations of machine learning models for predicting scientific replicability
Crockett, M.J., Bai, X., Kapoor, S., Messeri, L., Narayanan, A. (2023)
Proceedings of the National Academy of Sciences, Volume 120. Link» View PDF»