Innovations at the intersection of machine and human intelligence are enabling new tools for communication and aiding human cognition. At the same time, the tools can also lead to biased or adverse outcomes for users. We are still at the very early stages of developing reliable, interpretable, and inclusive AI systems that are "human aligned." The research focus of the Minds and Machines Lab is to understand key bottlenecks to building equitable AI systems and developing solutions to overcome them. Examples of our research projects include: evaluating how data quality affects algorithmic audits, using machine learning to improve behavioral nudges in cybersecurity settings, and understanding biases in foundational language models.
