Manuel Blum is a well-known theoretical computer scientist and a Turing award winner. But more interestingly, he has the highest number of students who have gone on to win a Turing award (Leonard Adleman, Shafi Goldwasser, Silvio Micali) in the whole computer science. The list of his students is amazing and even more so if we include the students of his students.
Can anyone comment on Manuel's supervisory style? What makes him so successful in training exceptional researchers? Anything that can help other supervisors be more successful in training exceptional researchers?