Marvin Minsky (1927-2016) was an American cognitive scientist and one of the founding figures of Artificial intelligence, who co-founded the MIT Artificial Intelligence Laboratory and helped organize the 1956 Dartmouth workshop. His 1969 book "Perceptrons," co-authored with Seymour Papert, is widely credited with curbing funding for neural-network research for over a decade.
Founding MIT's AI lab
Minsky studied mathematics at Harvard and Princeton, where his 1954 doctoral dissertation examined neural networks and reinforcement learning, work that included building SNARC, an early neural-network learning machine, in 1951. In 1959 he co-founded what became the MIT Artificial Intelligence Laboratory with John McCarthy, establishing MIT as one of the field's leading centers alongside Stanford and Carnegie Mellon.
"Perceptrons" and the first AI winter
In 1969 Minsky and Papert published "Perceptrons," a rigorous mathematical analysis of the single-layer Perceptron model pioneered by Frank Rosenblatt. The book proved that such networks could not compute certain simple functions, notably XOR, and its skeptical framing is often blamed, along with funding cuts following the UK's 1973 Lighthill report, for redirecting research and investment away from neural networks toward Symbolic AI approaches for roughly fifteen years, a period later remembered as part of the AI winter. Critics have argued the book's limitations were overstated or misapplied to multi-layer networks, which do not share the same constraints, but the chilling effect on funding was real.
"Society of Mind" and later work
Minsky's 1986 book "The Society of Mind" proposed that intelligence emerges not from any single mechanism but from the interaction of many simple, specialized processes, or "agents," working together, an idea influential in cognitive science and later multi-agent AI framings, including modern multi-agent systems. He continued writing about the nature of mind and machine consciousness, including "The Emotion Machine" (2006), and remained a vocal advocate for pursuing broad, human-level machine intelligence rather than narrow applications, a stance connected to later debates over Artificial general intelligence.
Legacy
Minsky received the Turing Award in 1969 and remained an influential, sometimes polarizing figure until his death in 2016. His symbolic and cognitive-architecture approach to AI, along with his early skepticism toward neural networks, stands in direct contrast to the Deep learning revolution led by researchers such as Geoffrey Hinton, whose backpropagation-trained multi-layer networks overcame the very limitations Minsky and Papert had identified in single-layer perceptrons.