NPR | Marvin Minsky, who pioneered artificial intelligence research, dies at age 88

host Robert Siegel of All Things Considered speaks to Ray Kurzweil
February 5, 2016

NPR’s Robert Siegel speaks to Ray Kurzweil, the inventor and futurist, and founder of Kurzweil Technologies, about Marvin Minsky, PhD.

Minsky, who was a founding father of artificial intelligence, has died at the age of 88. Kurzweil is a Director of Engineering at Google.

NPR | All Things Considered
episode title | Marvin Minsky who pioneered artificial intelligence research dies at the age of 88

full transcriptHost Robert Siegel talks with Ray Kurzweil

Robert Siegel — left                             Ray Kurzweil — right

Robert Siegel: A founding father of artificial intelligence has died at the age of 88.

— soundbite of archived recording —

Marvin Minsky: I’m Marvin Minsky, and I teach at MIT in the subjects of theories about how to make machines that are intelligent – whatever that means.

Robert Siegel: Marvin Minsky, PhD spent decades studying what that means. He was a winner of a Turing Award, the highest honor in computer science. He’s passed away at age 88.

That clip of him was from a video interview he gave to Ray Kurzweil — himself a computer scientist, writer, inventor and futurist — now a director of engineering at Google. Ray Kurzweil welcome to the program.

Ray Kurzweil: My pleasure.

Robert Siegel: How do you sum up Marvin Minsky’s contribution to computer science?

Photo of Marvin Minsky, PhD, who was a computer scientist in the field of artificial intelligence, born 1927 and passed away in 2016. Minsky’s many inventions include the first head mounted graphical display in 1963. He created the confocal microscope in 1957, a predecessor to today’s widely used confocal laser scanning microscope. His extensive background in math enabled him to pioneer the first days of the commercial computer. He held a math doctorate from Princeton University. He’s best known for his popular mainstream books on the possibility of computers achieving human level thinking. — credit | Royal Society of Chemistry

Ray Kurzweil: That’s really hard to do. The first thing I would say is he was the consummate educator. I mean, that was his great joy and passion. When I was 14 in 1962, I wrote him a letter and he invited me up and spent hours as if he had nothing else to do.

There’s an interesting anecdote with my daughter. We went out to dinner, and Marvin and Amy, who was then 11, were building a grand structure using all the silverware at this restaurant table, experimenting with different ways to make stable structures out of the tableware.

He just loved working with young people. But he was also a scientist, a mathematician, an inventor, an engineer, a roboticist, a writer, a philosopher, a polymath, a poet, an musician and a student of human nature and thinking.

And he invented, really, the two principle schools of thought in artificial intelligence — both the so-called symbolic school and connectionist schools. He really was the father of artificial intelligence.

Marvin Minsky, PhD co-founder of the Massachusetts Institute of Technology Artificial Intelligence Lab, did pioneering work in neural networks, knowledge representation, linguistics and robotics. This arm is one in a series of manipulators built at the lab. Photo was taken around 1970. — credit | Computer History Museum

Robert Siegel: Today, we’re not surprised by hearing that somebody has made a study of artificial intelligence or that they specialize in computer science. In the 1950s, Minsky’s work must have seemed prophetic to people.

Ray Kurzweil: Yeah, it was very prophetic. I mean, 1951, he actually built a neural net, which, you know, attempt to simulate how the brain works. And this was very, very early. He really was one of humanity’s great thinkers. He was also my only mentor, so it’s a great loss to me, but I greatly value the opportunities I’ve had.

Robert Siegel: When you say neural net, how would you define…

Ray Kurzweil: Well, it’s simulating how we believe the human brain works, which are sort of self-organizing modules of neurons. I wrote a book recently inspired by Minsky called How To Create A Mind, which describes this technique.

You let these sort of networks of neurons figure out on their own through practice and through learning how to solve a problem. And there’s actually been just recent breakthroughs that we can now go to many levels of neural nets and develop, you know, fairly abstract thinking.

Robert Siegel: In that same video interview that you did, Minsky says that machines in the future will just be programming themselves.

— soundbite of archived recording —

Ray Kurzweil: So we are part chimpanzee, and these future machines I think will be part human ’cause they’ll have been derived from us.

Marvin Misnky: They’ll be derived, but they might recompile themselves and say, well, we have all these things we inherited from humans and they make us so slow that let’s rewrite all the code and take out the old comments.

Ray Kurzweil: He was an optimist and felt that artificial intelligence would enhance the world, which is what I believe. Now, you know, technology is a double-edged sword and he was aware of that. He worked on ways to keep these technologies safe.

It’s somewhat of a shame that artificial intelligence is now coming into its own and we’re seeing dramatic advances just in the last few years and he’s not going to get to see the full fruition of what he worked on. His vision now is being realized, and we still have a long way to go. But he’s really accurately the father of artificial intelligence.

Robert Siegel: Ray Kurzweil, thanks for talking with us today.

Ray Kurzweil: My pleasure.

Robert Siegel: Ray Kurzweil was talking about the late Marvin Minsky, who died on Sunday night in Boston of a cerebral hemorrhage at age 88.

related reading:
NPR | shows: All Things Considered