Gary Marcus
Gary Marcus is a cognitive scientist, writer, and NYU emeritus scientist whose work examines rules, learning, and dependable AI. His MIT doctorate studied rules and exceptions in language. In The Algebraic Mind, he explored how symbol manipulation and neural networks could fit together. He later helped found Geometric Intelligence, acquired by Uber in 2016, and the robotics company Robust.AI. With Ernest Davis, he wrote Rebooting AI. His public work includes a 2018 critique of deep learning and Senate testimony calling for independent checks of AI systems. His recent response to an OpenAI mathematics announcement continues his concern with evidence and wider applicability. These critiques are his positions.
- Now at
- New York University
- Works
- 207
- Citations
- 10,117
- h-index
- 43
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
1993
Starting with language and rules
He completes an MIT doctorate on rules and exceptions in language.
-
2014–2016
Building a company around the problem
He helps found Geometric Intelligence; Uber acquires it in 2016.
-
January 2018
Putting the critique in writing
He publishes a critique arguing that general intelligence will require additional techniques.
-
September 2019
A book about dependable machines
He and Ernest Davis publish Rebooting AI, arguing for dependable systems.
-
October 2026
The same question meets a new result
The NYU emeritus scientist and writer asks for clearer evidence around a new mathematics result.
Sources and credits
Photo: Athena Vouloumanos, via Wikimedia Commons. License: CC BY 4.0. Source: commons.wikimedia.org. Cropped to a square and resized. Ask us to remove or replace this photo.
- NYU: Minds, Brains, and Machines faculty
- MIT: Marcus doctoral thesis, 1993
- MIT Press: The Algebraic Mind
- Turing Institute: Geometric Intelligence acquisition, 7 December 2016
- Deep Learning: A Critical Appraisal, 2018
- Publisher: Rebooting AI
- Robust.AI: Grace and Carter announcement
- Marcus: written Senate testimony, 16 May 2023
- Marcus: response to OpenAI mathematics announcement, 7 October 2026