Interesting side note: his friend Persi Diaconis (mentioned in the article) is the author of many interesting probability papers, including the proof that "7 riffle shuffles is enough" (https://statweb.stanford.edu/~cgates/PERSI/papers/aldous86.p...)