Berkeley — IEOR, the Dreyfus brothers, and Hinton's paper

You'll note that my degree at UC Berkeley was Industrial Engineering & Operations Research. We were on the third floor of Etechevary Hall, just north of Hearst Street. Soda Hall is where our parking lot was. The floor below us was the Nuclear Engineers, and the floor above was the Civil Engineers. They both called us 'Imaginary Engineers', because we designed systems, not things, or even just simulated things - my focus was simulation and modeling of dynamic systems. My primary advisors were C. Roger Glassey and Stuart Dreyfus. Stuart I knew from childhood, since he and his brother Hubert Dreyfus were both members of my mom's Hindu Buddhist Ashram, which we called Ramagiri, but was officially known as the Blue Mountain Center for Meditation. My mother worked for Hubert Dreyfus in building T-8 on campus, immediately to the West of Evan's Hall, which was the computer science building. Hubert and Stuart Dreyfus co-taught a course combining philosophy and engineering, trying to apply rigor and logic to philosophy. I spent most afternoons in various buildings on campus - I had the run of the place. I had been on campus since I was five years old - mom brought me to a lecture by Jane Goodall, where they showed a movie about primate behavior in africa - chimps.

I used to go to Evan's Hall with Jim Michael, my mothers boyfriend for three years, and my surrogate father. He was rich, from Piedmont, and his family owned Grodin's, a high end menswear chain in the bay area. He is the one that introduced my mother to Ramagiri / Blue Mountain Center of Meditation. He taught me how to use punch card machines, program with them, feed them into the mainframe, and then read the printer output to see if they worked. It was cool but very monotonous at the same time. I wondered where the computers from 2001 Space Odyssey were - that was my idea for a computer. But I learned.

Stuart Dreyfus and C. Roger Glassey did seminal work on the mathematics of neural nets and back-propagation. They gave me Geoffrey Hinton's paper on back-propagation, and showed me a neural net running on Roger's Next machine. (Berkeley, Stanford and Carnegie Mellon were the academic institutions that backed Steve Jobs and Next.)