John Hopfield

Hopfield has written that he was born in Chicago in 1933 during the World’s Fair, where his father had set up the physics exhibit; both his parents were physicists. As a child he liked taking apart and repairing household objects, bicycles and radios.⁠[1]

He earned his undergraduate degree in physics from Swarthmore College in 1954.⁠[2]

After receiving his Cornell PhD in 1958, whose thesis studied the interaction of light with excitons in solids, he spent two years in the theoretical group at Bell Labs and began teaching at the University of California, BerkeleyUniversity of California, BerkeleyThe University of California, Berkeley (UC Berkeley) is a campus of the University of California system; the University of California was founded in 1868 through a merger of two institutions and moved to Berkeley, California, in 1873.Open the full entry in 1961.⁠[3]

His Cornell adviser was Albert Overhauser, whom he credits with giving him “ownership of an interesting question”; the thesis led to the polariton concept.⁠[1]

In 1964 he joined Princeton University as a professor of physics, serving until 1980.⁠[4]

In 1968–1969 he held a Guggenheim Fellowship for theoretical condensed matter physics at the Cavendish Laboratory of the University of Cambridge.⁠[5]

He has been a member of the US National Academy of Sciences since 1973.⁠[6]

In 1974 he described kinetic proofreading, a mechanism for error correction in biological reactions such as protein synthesis.⁠[2]

In 1980 he moved to the California Institute of Technology as professor of chemistry and biology, staying until 1996.⁠[7]

On April 15, 1982, he published “Neural networks and physical systems with emergent collective computational abilities” in the Proceedings of the National Academy of Sciences, showing how a content-addressable memory that recovers an entire memory from a sufficiently large fragment can emerge from systems of many simple neurons.⁠[8]

This associative neural network later became known as the Hopfield network and was one of the early ideas that catalyzed the development of modern deep learning.⁠[9]

In 1985 he and D. W. Tank published a paper in Biological Cybernetics showing that highly interconnected networks of nonlinear analog neurons are extremely effective at computing decisions in optimization problems.⁠[10]

With Richard FeynmanRichard FeynmanRichard Feynman (Richard Phillips Feynman, 1918–1988) was an American theoretical physicist who rebuilt quantum electrodynamics with the path-integral formulation and Feynman diagrams, for which he shared the 1965 Nobel Prize in Physics with Julian Schwinger and Sin-Itiro Tomonaga. He took part in the Manhattan Project during World War II and then taught for decades at the California Institute of Technology, where his introductory physics course became The Feynman Lectures on Physics; in 1986, as a member of the presidential commission investigating the space shuttle Challenger accident, he used ice water to demonstrate that the O-ring lost its resilience at low temperature.Open the full entry and Carver Mead he co-taught a yearlong course, “The Physics of Computation,” which three years later evolved into an interdivisional program; in 1986 he co-founded Caltech’s Computation and Neural Systems.⁠[7]

He returned to Princeton in 1997 with an appointment in molecular biology and became emeritus as the Howard A. Prior Professor of the Life Sciences in 2008.⁠[4]

He was instrumental in creating the Princeton Neuroscience Institute.⁠[11]

On January 1, 2005, he succeeded John Bahcall as vice president of the American Physical Society, becoming president-elect in 2006 and president in 2007; at the time he also chaired Princeton’s molecular biology department.⁠[3]

In 2019 the Franklin Institute awarded him the Benjamin Franklin Medal in Physics for applying theoretical physics concepts to biological questions including neuroscience and genetics, with significant impact on machine learning.⁠[2]

On February 29, 2020, he appeared on episode 76 of the Lex FridmanLex FridmanLex Fridman is a Soviet-born, US-raised computer scientist and podcast host who works as an MIT research scientist on human-AI interaction and autonomous driving. His Lex Fridman Podcast is known for hours-long conversations with guests including heads of state and technology leaders.Open the full entry Podcast, discussing differences between biological and artificial neural networks, a physics view of the mind, Hopfield networks and associative memory, Boltzmann machines, consciousness, attractor networks, brain-computer interfaces and the meaning of life.⁠[9]

In April 2023 he and Frank Jülicher received the first IUPAP Medal for the Physics of Life from the International Union of Pure and Applied Physics, recognizing his pioneering contributions on electron transfer in proteins, proofreading in molecular information transmission and the dynamics of neural networks.⁠[6]

On October 8, 2024, he and Geoffrey Hinton won the Nobel Prize in Physics, carrying 11 million Swedish kronor, for discoveries that “enabled machine learning with artificial neural networks”; Ellen Moons, chair of the Nobel Committee for Physics, said their discoveries and inventions form the building blocks of machine learning.⁠[12]

Aged 91 and visiting England, he joined Princeton’s press conference by Zoom; on artificial intelligence he said, “I don’t know. I worry.”⁠[4]

At the press conference he said that “physics is trying to understand how systems work” and that new physics will be found in understanding how the brain operates.⁠[11]

He is the Howard A. Prior Professor in the Life Sciences Emeritus and Professor Emeritus of Molecular Biology at Princeton University.⁠[11]

Princeton University: Official website: https://www.princeton.edu/ (opens in a new window)⁠[4]

California Institute of Technology: Official website: https://www.caltech.edu/ (opens in a new window)⁠[7]

University of California, Berkeley: Official website: https://www.berkeley.edu/ (opens in a new window)⁠[6]