Assessing Diversity in AI Research

This chapter explores the measurement and significance of diversity in AI research, highlighting the use of topic modeling to assess the distribution of ideas over time. It discusses the impact of the private sector's dominance on research diversity, revealing a troubling trend of stagnation in variety despite an increase in publications. The chapter also delves into methodological approaches used to analyze AI literature across organizations, underscoring the disparities in thematic diversity between private companies and academic institutions.