Research

Key Considerations developed with a lens for adoption by U.S. government departments and agencies critical to national security. However, they are relevant more generally for the design, construction, and use of AI systems.

Co-drafted a white paper on pandemic response and preparedness, co-authored by NSCAI Commissioners Dr. Eric Horvitz, Hon. Mignon Clyburn, Dr. José-Marie Griffiths, and Dr. Jason Matheny.

Forecasts of job loss due to automation vary widely - differing by tens of millions of jobs. This report, commissioned by Keystone Research Center, adds precision to forecasts for the trucking industry. Produced with economists, sociologists and ML experts at Carnegie Mellon University and University of Pennsylvania, it projects geo-based job and wage loss across America. After, it scopes a wage insurance policy solution to cover displaced workers.

This project, advised by machine learning academics, critiques a model used to predict recidivism and produces alternative models. Included are classification matrices that assess and report on disparate performance across gender, age, and race.

Disinformation and hate speech spread online have led to violence offline - in countries as diverse as India, Myanmar and the US. This research explores common threads across these cases to identify shortcomings of policies employed to date, to propose solutions for content moderation, and to form U/X recommendations for reporting abusive content.

Psychological shortcuts can cause people to fall prey to disinformation - or they can be part of the solution. This paper explores prior policies for tackling disinformation and the cognitive biases that have undermined their efficacy. It then outlines content moderation approaches that play off these biases while helping users critically assess content validity.

This paper shares comparative research, benchmarking and policy recommendations for Pennsylvania’s Department of Human Services to enhance economic and education opportunity for its clients. The analysis is segmented to produce demographic- and needs-based insights for better allocating services and targeting interventions.

Industry and political leaders alike are calling for solutions to curb unprecedented income inequality in the age of AI - from Universal Basic Income (UBI) to a tax on AI. This essay explores a policy instrument that gets less media coverage despite its merits, including hints of precedent for political feasibility.