During this time in history when the world is undergoing great and uncertain change, it is worth reconsidering the relationships connecting computer science, social science, and the humanities. One popular form of artifcial social intelligence, namely recommender systems, can become a far more valuable tool for research on the arts, beginning with movies and computer games, then extending to all the other art forms. Survey research using questionnaires is well established, but social media offer great improvements in both methodology and application. It is time also for a revival of computer simulation as a tool for development of rigorous theories of social interaction, both abstractly and in human experience of virtual worlds. The study of literature is a major focus of education, yet it needs creative modernization, both to make use of computer collection and analysis of data, and to include the vast new online medium of millions of works of amateur fiction. Industry and an enthusiastic social movement have oversold pure artificial intelligence, as for example illustrated by AI's failure to understand poetry, and we must develop the best methods for human use of the new technology. Together, these developments offer opportunities and challenges for both industry and gov-ernment policy. Unless arti?cial intelligence converges successfully with the social sciences and humanities, it will be justly accused of insincerity, exaggeration, extremely negative impacts, and subservience to the inhumane ambitions of technocrats and other societal elites.