Hyped or Transformational? AI in Higher Education

Abstract

This paper explores the difference between the transformation idea and the reality of how colleges and universities leverage technologies for teaching, learning and assessment. We seek to understand why a structural and operational transformation of these institutions rarely happens and offer an understanding of the structural, operational, financial, and political constraints that prevent transformation but encourage continuous improvement. More specifically, we suggest that AI will not deliver on its promise to finally “solve” the so-called “two sigma” problem, which arises from the work of Bloom (1984), who showed that students taught through tutoring performed better by two standard deviations than those taught in a traditional classroom.