معرفی
Joseph Jay Williams is an Assistant Professor in Computer Science at the University of Toronto, with courtesy appointments in Psychology, Statistics, Industrial Engineering, Economics, and the Vector Institute for Artificial Intelligence. He leads the Intelligent Adaptive Interventions Lab, focusing on transforming digital components into adaptive micro-laboratories that use machine learning algorithms to optimize behavior change interventions in education and health.
- PhD in Computational Cognitive Science from UC Berkeley
- Former Assistant Professor at National University of Singapore
- Ex-Research Fellow at Harvard's VPAL and postdoc at Stanford
His research spans four themes: (1) Adaptive Experimentation Tools for education and collective intelligence, (2) Human-AI Co-Design with large language models, (3) Personalization Algorithms for mental/physical health interventions, and (4) Statistically Sensitive Algorithms that balance reward optimization with rigorous inference. The lab's AdaptEx framework has impacted over 500,000 users and won a $1M Xprize.
Recent work includes ABScribe for human-AI co-writing experiments and LLM-driven prompt engineering frameworks. The lab develops MOOClet technology that enables dynamic A/B testing of educational components like explanations, hints, and emails. Algorithms prioritize both participant benefit and statistical power - for example, Thompson Sampling variants that contextualize interventions based on user state while maintaining scientific validity.
- $1M Xprize for AI-driven education
- $3M NSF Cyberinfrastructure grant
- Best Paper Awards at CHI 2023 and HCOMP 2023
- DARPA AI Tools Competition Finalist
- $925K ONR grants (2018-2024) for personalization research
Current projects integrate adaptive experimentation with chemical/materials science through the Acceleration Consortium's $200M federal initiative. The lab maintains collaborations with institutions like Northwestern Medical School, UCSF Medical School, and Worcester Polytechnic Institute (co-PI on NSF grant). The research vision targets Nobel-Prize-level integration of empirical sciences, machine learning, and statistics to accelerate discoveries across disciplines.
