Rodney Brooks (born December 30, 1954, in Adelaide, Australia) is a roboticist and computer scientist whose research on behavior-based Robotics reshaped how mobile robots are built and who later became a prominent skeptic of near-term Artificial general intelligence predictions. He earned a PhD in computer science from Stanford University and went on to direct the MIT Computer Science and Artificial Intelligence Laboratory.
Subsumption architecture
In the 1980s, working against the dominant paradigm of centralized, symbolic planning in AI research (see Symbolic AI), Brooks developed the subsumption architecture, a layered control system in which simple behaviors act directly on sensor input without building an internal symbolic model of the world. His influential papers, including "Elephants Don't Play Chess" and "Intelligence Without Representation," argued that robust real-world behavior could emerge from many simple, reactive layers rather than from explicit reasoning, a position that significantly influenced the later development of Embodied AI as a research direction distinct from purely software-based machine learning.
iRobot and commercial robotics
Brooks co-founded iRobot in 1990, a company that produced military and industrial robots before achieving mass-market success with the Roomba autonomous vacuum cleaner, first released in 2002 and grounded directly in subsumption-style reactive control. He later founded Rethink Robotics in 2008, which developed collaborative factory robots designed to work safely alongside humans, and subsequently founded a company working on warehouse robotics.
Criticism of AI predictions
From the 2010s onward, Brooks became known for publicly critiquing what he viewed as recurring overoptimism in predictions about AI and robotics timelines, including forecasts made by figures such as Ray Kurzweil. He maintains a running, dated scorecard of predictions, including his own, about self-driving cars, Humanoid robot deployment, and Artificial general intelligence, revisiting them publicly to check accuracy against actual progress. His central argument holds that commentators, including many inside the AI industry, systematically underestimate the difficulty of embodied, real-world tasks while overestimating the near-term significance of progress in purely digital domains such as Large language model benchmarks.
Reception
Brooks is widely respected within robotics for both his technical contributions and his willingness to publicly push back against AI hype cycles, a stance that has put him at odds with some industry narratives about imminent transformative AI. Supporters view his prediction-tracking as a valuable corrective in a field prone to overclaiming; critics counter that his skepticism about rapid progress in language and reasoning models, expressed in earlier writing, underestimated the pace of advances following systems such as GPT-4 and successor Reasoning model systems.