AI’s social sciences deficit

To create less harmful technologies and ignite positive social change, AI engineers need to enlist ideas and expertise from a broad range of social science disciplines, including those embracing qualitative methods, say Mona Sloane and Emanuel Moss in a comment piece in Nature.

Technologists are increasingly looking to social scientists to help fix the problem of harmful or biased AI through a focus on ethics or safety, or to develop AI that is aligned with human values. But they would do better by making full use of the social sciences, not just embracing quantitative methods. […] We argue that it is misguided to only frame social issues in quantitative ways, where they can be studied and measured in abstract lab situations or thought experiments: AI does not fail people in a lab; it fails them in real life, with real consequences.