Generate a tailored SOP for Dr. John Thickstun. Improve your application with a focused, well-structured draft.
John Thickstun is an Assistant Professor in the Department of Computer Science. His primary research interests lie in machine learning generative models, with a focus on methods for controlling behavioral models and their applications to various task-oriented contexts. John also explores the implications of generative models as a policymaking tool to regulate their outputs. His contributions to the field include advancing applications of generative models across multiple modalities including text, image, and music technologies. Previously, he was a Postdoctoral Scholar at Stanford University under the guidance of Percy Liang. John earned his PhD while co-advised by Sham Kakade and Zaid Harchaoui and completed his undergraduate studies in Applied Mathematics under Eugene Charniak. His work on the MusicNet dataset has significantly contributed to its transition to permanent hosting at Zenodo. He has been involved in workshops aimed at examining the intersection of AI and creative arts, such as his recent participation in a retrospective conversation about the future of storytelling. John is actively publishing blog posts regarding his research outcomes and public interest in generative music.
Stanford University • Stanford, CA
Conducted research in machine learning generative models, working with Percy Liang.
Stanford University • Stanford, CA
Teaching and conducting research in machine learning and generative technologies.
The Computer Science department emphasizes research potential. GRE General is currently optional but recommended for some tracks.