In experiments at six public universities, students assigned randomly to statistics courses that relied heavily on “machine-guided learning” software — with reduced face time with instructors — did just as well, in less time, as their counterparts in traditional, instructor-centric versions of the courses. This largely held true regardless of the race, gender, age, enrollment status and family background of the students.
The issued here is a very basic one. A teacher or professor with any reasonable sized class can't tailor the class to one student. Inevitably, the teacher must go on at greater length for some students on some topics, leaving others bored reviewing what they already know. A computer can provided individualized attention in the same way that a personal tutor does, and hence, can avoid this problem.