Marvin Minsky

AI Pioneer & Co-founder of the MIT AI Lab

Portrait painted after an image by Sethwoodworth · CC BY 3.0

About

Marvin Minsky (1927–2016) was an American cognitive scientist, born in New York City, who helped found artificial intelligence as a field. After a mathematics degree at Harvard (1950) and a Princeton PhD (1954) on neural-analog reinforcement systems, he co-signed the 1955 proposal for the Dartmouth workshop that gave the field its name, and in 1959 began with John McCarthy the project that became MIT's AI Laboratory; he stayed at MIT for the rest of his life. His 1951 SNARC is often considered the first artificial neural network, yet Perceptrons (1969), written with Seymour Papert, proved the limits of single-layer networks and is often blamed for the decline in neural-network research that followed. In The Society of Mind (1986) and The Emotion Machine (2006) he described a mind as a society of many interacting agents rather than a single unified thinker; he received the Turing Award in 1969 and advised Stanley Kubrick on 2001: A Space Odyssey. He took a $100,000 research grant from Jeffrey Epstein in 2002 and organized symposia on Epstein's private island in 2002 and 2011, the second after Epstein had registered as a sex offender.

Key Contributions

Co-signed the 1955 Dartmouth proposal that coined the term 'artificial intelligence'
Co-founded with John McCarthy the MIT project that became the Artificial Intelligence Laboratory
Built SNARC (1951), often considered the first artificial neural network, and later co-wrote Perceptrons (1969) with Seymour Papert on the limits of single-layer networks
Proposed in The Society of Mind (1986) that a mind is built from many simple interacting agents, extended in The Emotion Machine (2006)
Received the 1969 Turing Award; also invented the confocal microscope (1957) and advised on Kubrick's 2001: A Space Odyssey

Connections

John McCarthy

Collaborated

Computer Scientist & Father of AI

In 1955 Minsky, then a Harvard junior fellow, co-signed with McCarthy, Rochester and Shannon the Dartmouth proposal that named artificial intelligence and wagered that 'every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.' In 1959 the two began the MIT project that became its AI Laboratory. That sentence is still the field's working bet, and still its most argued-over one.

en.wikipedia.org · en.wikipedia.org

Ray Kurzweil

Influenced

Inventor, Futurist & Author

At fourteen Kurzweil wrote Minsky a letter; Minsky invited him to MIT and spent hours with him, and Kurzweil went on to study there with him, graduating in 1970. Remembering him in 2016, Kurzweil called Minsky 'my only mentor' and credited him as a pioneer of both the symbolic and the connectionist schools. The teacher spent his life on how a mind is organized; the student staked his name on when one would be built.

en.wikipedia.org · edge.org

Jaron Lanier

Influenced

Computer Scientist, VR Pioneer & Tech Philosopher

Lanier was never formally Minsky's student, yet calls him his most important mentor, and in his 2016 remembrance asked: 'Why was Marvin so generous to me? I gave him grief. I disagreed with him at every turn.' He credits Minsky's account of emotion and intelligence with displacing Freud's in how people now picture the mind. That one of the sharpest critics of AI-as-autonomous-being was formed by one of the field's founders is perhaps the best evidence that a mentorship can be made of disagreement.

edge.org · uctechnews.ucop.edu

Geoffrey Hinton

In contrast

AI Pioneer & Researcher

Perceptrons (1969), by Minsky and Seymour Papert, proved that a single-layer perceptron cannot compute functions like parity, and the field took it as a verdict on neural networks as such; the book is often blamed for the decline of neural-net research through the 1970s. Hinton spent the following decades on exactly the multilayer networks it was taken to have ruled out, and the revival he helped lead is the one Minsky and Papert answered in their 1988 epilogue. Read together they mark the field's hinge: a mind as a society of designed parts, or a mind as weights learned from data.

en.wikipedia.org · en.wikipedia.org