Austin van Loon
Sociologist · Computational Social Scientist
Class of 1956 Career Development Assistant Professor,
MIT Sloan School of Management
I study how people make sense of their social worlds — and how that shapes whether we learn from disagreement, coordinate, and act together. I also build methods that use AI to make social science itself more rigorous, including experiments that combine human participants with AI predictions.
Making Sense of Disagreement
In settings marked by disagreement and ambiguity, what happens next depends on interpretation: what information means, whose reading wins out, and how shared understandings turn into action. I study this in contexts ranging from political polarization and immigration attitudes to the design of social media platforms and the arrival of AI in organizations and higher education.
Social Science in the Age of AI
AI is transforming how social science gets done. Treating AI simulations as substitutes for people can produce confidently wrong conclusions — so I build methods that put AI to work without giving up rigor: experimental designs that mix human subjects with LLM predictions and stay unbiased no matter how wrong the model is, and frameworks for understanding what automated text analysis actually measures.
About
I am a sociologist and computational social scientist at the MIT Sloan School of Management, where I am the Class of 1956 Career Development Assistant Professor of Work and Organization Studies. I am a standing faculty member of the Organization Studies group, a core faculty member of the Economic Sociology program, and a member of the Managerial Communications group. I am also a member of the MIT Behavioral Research Lab Advisory Board.
Before MIT, I completed my PhD in Sociology at Stanford University, where my dissertation examined how artificial intelligence can be applied to the sociological study of meaning. I then spent a year as a postdoctoral associate at Duke University's Polarization Lab. I received my B.S. in Sociology and my B.A. in Psychology from the University of Iowa.
My work has appeared in outlets including Management Science, Sociological Methods & Research, PNAS, Nature Human Behaviour, Nature Computational Science, Scientific Reports, Communications Psychology, Social Science Research, and American Behavioral Scientist.
Recent highlights
- July 2026“Designing social media to promote productive political dialogue” published in Scientific Reports.
- July 2026
mixedsubjects, our R package for designing and analyzing experiments that combine human subjects with LLM predictions, released on CRAN (v1.0.0). - 2026Co-organizing the Junior Faculty in Organizational Theory Conference, held at MIT Sloan for the first time; invited talks at the Yale Computational Social Science Seminar and Northwestern's symposium on validating generative-AI-based social science.
- August 2025Invited News & Views piece, “Experimenting on AI can (sometimes) teach us about ourselves,” published in Nature Computational Science.
- 2025“The Mixed Subjects Design” published in Sociological Methods & Research.
- 2025Invited talks at NYU (PRIISM), SICSS-Paris, and the MIT Conference on Digital Experimentation.
Teaching
This year at MIT Sloan I teach AI Foundations (required AI-literacy sessions for MBA and EMBA students), lead AI Builder Space (a new course where students build products, ventures, and productivity tools with AI), and teach the PhD course Designing Empirical Research in the Social Sciences. I previously taught the core MBA course Communication for Leaders and co-created the Generative AI Hackathon for Social Good. Details, software, and open materials live on the Software & Teaching page →
Contact
Email: vanloon@mit.edu
Office: E62-343, MIT Sloan School of Management, Cambridge, MA
Get in touch
- Researchers: I'm glad to hear from people interested in mixed-subjects designs, the projects above, or potential collaborations.
- Prospective PhD students: I work with doctoral students in MIT Sloan's PhD program (Organization Studies and Economic Sociology) as well as MIT's Institute for Data, Systems, and Society (IDSS). You don't need to email before applying — but you're welcome to.
- Journalists & practitioners: I'm happy to talk about AI and social science, political division, and designing healthier online platforms.