Remembering Marvin Minsky
The world lost an incredible man this week with the passing of Marvin Minsky. Those who know him or know of him usually know his role as a father of the field of artificial intelligence. Through his writing and through his teaching — often one and the same — he had a far-reaching impact on the world of AI today.
While there are many people who knew Marvin better than I or spent much more time with him, he was one of my strong influences for the last twenty plus years. I bought my first dog-eared used copy of Society of Mind sometime in the early 1990s, not all that many years after it was first published. I remember reading it at least a couple of times then before I finished high school, but the thing that stands out the most in my memory is how it just made sense to me. His ideas about creating an intelligent, thinking “mind” through many smaller, less complex components still directly influence the way that I architect intelligent systems to this day.
When I arrived at the MIT Media Lab 15 years ago, I was amazed by the people who were there whose books I had read, but who I had never imagined meeting. Nicholas Negroponte, Sherry Turkle, and of course Marvin Minsky. People whose words and thoughts had inspired me for years who I never thought I would meet, let alone become friends with some of them.
I remember in my second semester as a young grad student at MIT wanting to take Marvin’s evening seminar, also called Society of Mind. That first night we all gathered in the basement auditorium of the Media Lab and filled at least a hundred seats. He gave his introductory lecture and then explained the topic of the essay that each of us was to write if we wanted to be chosen as one of the twenty or so students to actually spend one evening a week the rest of the semester in his course. Through thinking about and writing that essay over the next week, I first got to know Push Singh, who was the TA for the class and the reader (and therefore decision-maker) of those essays. (Another great in AI who I looked up to and who we have lost.)
I was fortunate to make the cut and spend Tuesday evenings (I think it was) for the rest of the semester in a small classroom in the basement of the Media Lab with Marvin. We, the students, would have done the required reading — some from Society of Mind, but much that semester from very early drafts of The Emotion Machine that he was starting to write at the time — and then Marvin would walk in and take his seat at the front. “So what do you want to talk about tonight?” he would say in his inimitable style. And that was the format of the seminar nearly every week. Marvin would ask us what was on our minds, presumably related to the readings, and we would launch into a discussion.
Marvin and I were definitely on opposite sides of a big topic at that point in time. This was during his period of loudly proclaiming that building robots was a waste of graduate students’ lives. Yet I had just arrived at MIT to undertake the building of robots. Or more importantly to study what happens when people interact with them, but that required actually building them at that time. I have to hope that with what has become of that work over the past fifteen years that Marvin would be proud of the results of those hundreds of hours toiling in the labs and shops to build many robots.
While I was never particularly close to Marvin during my time at MIT, his writings, discussions, and thinking have had a profound influence on my life and work. I believe that what I’m building today at Catalia Health would not exist if it weren’t for a number of people who I have learned from over the past couple of decades and Marvin Minsky was definitely one of those people.
Rest in peace, Marvin. We miss you.