Tomaso Poggio

Tomaso A. Poggio was born on September 11, 1947, in Genoa, Italy. By his account he attended the Jesuit school Arecco from age six to eighteen, then studied physics at the University of Genoa.⁠[1]

In 1971 he received his doctorate in theoretical physics from the University of Genoa.⁠[2]

He recalled that Werner Reichardt hired him as an assistant after a visit to the Max Planck Institut für Biologische Kybernetik, and he arrived in Tübingen in September 1971; in June 1972 he married Barbara Venturini Guerrini.⁠[1]

There, working with W. Reichardt, he quantitatively characterized the visuo-motor control system of the fly.⁠[3]

In June 1976 he and D. Marr issued the MIT AI Lab memo “Cooperative Computation of Stereo Disparity” (AIM-364), which analyzed the correspondence problem between two images and derived a cooperative algorithm that extracts disparity information from random-dot stereograms.⁠[4]

In 1981 he left Tübingen to become an associate professor at MIT.⁠[2]

He was a Corporate Fellow of the supercomputer company Thinking MachinesThinking Machines LabThinking Machines Lab is an artificial intelligence research and product company founded by former OpenAI chief technology officer Mira Murati and several former OpenAI researchers, with the aim of making AI systems more understandable, customizable and collaborative. In July 2025 it closed a $2 billion seed round at a $12 billion valuation; its main product is Tinker, an API that lets researchers and developers fine-tune open-weight models, and since 2026 it has also released the open-weights Inkling model family.Open the full entry Corporation, later served as a director of PHZ Capital Partners and of Mobileye, and was involved in starting or investing in companies including Arris Pharmaceutical, nFX, Imagen, Digital Persona and DeepMindGoogle DeepMindGoogle DeepMind is Google’s artificial intelligence research organization, formed in April 2023 by merging DeepMind, which Google acquired in 2014, with the Brain team from Google Research, and led by CEO Demis Hassabis. It developed AlphaGo, AlphaFold, and the Gemini family of models, and Hassabis and John Jumper received the 2024 Nobel Prize in Chemistry for AlphaFold’s protein structure prediction work.Open the full entry.⁠[5]

In a 1990 NIPS paper with Federico Girosi and Bruno Caprile, he framed learning an input–output mapping from examples as approximating a multidimensional function, noting that he and Girosi had earlier shown the equivalence between regularization and a class of three-layer networks they called “regularization networks.”⁠[6]

In 1999 he and Maximilian Riesenhuber published “Hierarchical models of object recognition in cortex” in Nature Neuroscience, describing a hierarchical object-recognition model consistent with physiological data from inferotemporal cortex and based on a MAX-like operation.⁠[7]

Naturalized in 1994, he joined the faculty of MIT’s McGovern Institute for Brain Research in 2000 and that year received an honorary Laurea from the University of Pavia for the Volta Bicentennial.⁠[2]

In October 2009, while director of the Center for Biological and Computational Learning, he received the Okawa Prize from the Okawa Foundation for Information and Telecommunications, its 13th U.S. winner; the citation honored his “outstanding contributions to the establishment of computational neuroscience.”⁠[8]

On September 9, 2013, the National Science Foundation announced funding for the MIT-based Center for Brains, Minds and Machines (CBMM), headed by Poggio, initially $25 million over five years; the center grew out of the MIT Intelligence Initiative, which he and Joshua Tenenbaum proposed and which launched in 2011.⁠[9]

He said the early dreams of artificial intelligence remained unfulfilled and “it’s time to try again”; he said the center would study intelligence in terms of algorithms and strategies, its implementation in neural circuits, its development in children, and social intelligence.⁠[10]

In 2014 the Society for Neuroscience awarded him the $25,000 Swartz Prize for Theoretical and Computational Neuroscience, presented at Neuroscience 2014 in Washington, D.C.; SfN president Carol Mason called him a “founder and driving force” in computational neuroscience.⁠[11]

In March 2017 he and co-authors including Hrushikesh Mhaskar published a review in the International Journal of Automation and Computing on the conditions under which deep networks can be exponentially better than shallow ones and avoid the curse of dimensionality.⁠[12]

That year the IEEE Computer Society Technical Committee on Pattern Analysis and Machine Intelligence gave him the PAMI Azriel Rosenfeld Lifetime Achievement Award.⁠[13]

On June 11, 2018, MIT announced the leadership of its Quest for Intelligence, naming him founding scientific advisor of “The Core,” directed by James DiCarlo with Josh Tenenbaum and Leslie KaelblingLeslie KaelblingLeslie Pack Kaelbling is an American computer scientist and roboticist and the Panasonic Professor of Computer Science and Engineering at MIT, known for research on reinforcement learning, planning with partially observable Markov decision processes and robot task and motion planning, and for founding the open-access Journal of Machine Learning Research.Open the full entry as scientific directors.⁠[14]

The Quest’s website recalls that CBMM, funded long-term by NSF in 2013 and led by Poggio, set out to understand how the brain makes the mind, how brain circuits learn and compute, and how intelligence might be replicated in machines, paving the way for the Quest.⁠[15]

On January 19, 2019, he appeared on episode 13 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, “Brains, Minds, and Machines”; the episode description said his work had been cited over 100,000 times and noted that he had advised researchers and entrepreneurs including Demis HassabisDemis HassabisDemis Hassabis is a British artificial intelligence researcher and entrepreneur, co-founder of DeepMind and its CEO and that of Google DeepMind from 2010 to 2026, and since then Chair of Google DeepMind, Chief Scientist of Alphabet and CEO of Isomorphic Labs; a former chess prodigy and game designer, he shared the 2024 Nobel Prize in Chemistry with John Jumper for AlphaFold’s protein structure prediction, and has appeared twice on the Lex Fridman Podcast.Open the full entry, Amnon Shashua and Christof Koch.⁠[16]

He, H. Kuehne, H. Jhuang, E. Garrote and T. Serre received the 2021 Helmholtz Prize for the paper “HMDB: A large video database for human motion recognition”; the prize is given by TCPAMI at ICCV for fundamental contributions to computer vision.⁠[17]

Having won the first edition of the international scientific award “Ratio et Spes,” he received the Kampé de Fériet award in 2022.⁠[5]

In January 2026 he discussed learning theory with Paul Middlebrooks on The Transmitter’s “Brain Inspired” podcast, explaining principles such as “compositional sparsity,” the idea that many simple functions connected in hierarchies can perform complex functions.⁠[18]

As of October 2026 the McGovern Institute lists him as Eugene McDermott Professor Emeritus and former CBMM director, with research focused on the mathematics of deep learning and the computational neuroscience of the visual cortex.⁠[19]

His lab website says he mentored figures such as Christof Koch, Amnon Shashua and Demis Hassabis, and states that as of 2024 he was the largest individual recipient of NSF AI funding over a 14-year period.⁠[20]

He also serves as a scientific advisor to the MIT Siegel Family Quest for Intelligence.⁠[21]

Personal website: Public link: https://poggio-lab.mit.edu/ (opens in a new window)⁠[20]

MIT: Official website: https://www.mit.edu/ (opens in a new window)⁠[5]

McGovern Institute for Brain Research: Official website: https://mcgovern.mit.edu/ (opens in a new window)⁠[11]

Center for Brains, Minds and Machines: Official website: https://cbmm.mit.edu/ (opens in a new window)⁠[17]