Human Computation

The real-world utility of machine learning due to improvements in computing speed have inspired a new generation of AI researchers, who continue to push the envelope on human-like capabilities that can be fully automated.  Nonetheless, superhuman capabilities remain within the purview of two strategies: those that leverage the complementary abilities of humans and machines, and those that effectively integrate the cognitive output of many people.  Indeed, new “Crowd AI” systems, which combine those approaches, have led to unprecedented capabilities, and for certain problem classes may constitute the only viable solutions today. At the same time, growing awareness of the broad applicability of AI to societal problems including humanitarian issues has led to recent initiatives in the public and private spheres, including Microsoft’s AI for Good, Google’s AI for Social Good, as well as the UN’s AI for Good Global Summit, which seeks to apply machine intelligence to 17 sustainable development goals set by the UN General Assembly. Thus, this special issue reports on initiatives at the intersection of these two movements by examining the opportunities, methods, and impacts of Crowd AI as applied to societal problems.