This year’s Turing Award, an honor often dubbed the “Nobel Prize of Computing,” goes to two A.I. researchers who laid the foundations for tech breakthroughs like OpenAI’s GPT. Andrew Barto, a researcher at the University of Massachusetts Amherst, and Richard Sutton, a professor at Canada’s University of Alberta, will share the $1 million prize, as announced today (March 5).
Named for the British mathematician Alan Turing, the award bestows a cash prize with financial support from Google (GOOGL) and is given annually by the Association for Computing Machinery (ACM). Past winners included A.I. researchers like Geoffrey Hinton, Yoshua Bengio and Yann LeCun, who received the 2018 Turing Award for their work in artificial neural networks.
Barto and Sutton won this year’s award for their contributions to the field of reinforcement learning, a process used to improve the behavior of machines. “Their work has been a lynchpin of progress in A.I. over the last several decades,” said Jeff Dean, Google’s chief scientist, in a statement. “The tools they developed remain a central pillar of the A.I. boom and have rendered major advances, attracted legions of young researchers, and driven billions of dollars in investments.”
The duo began working together in 1978 at UMass Amherst, where Barto served as Sutton’s Ph.D. and postdoctoral advisor. In the following years, they collaborated on numerous papers that shaped key algorithms and techniques of reinforcement learning and published the 1998 textbook Reinforcement Learning: An Introduction, which has been cited more than 75,000 times and remains the field’s standard reference.
While their work took place decades ago, it remains more relevant than ever, noted Yannis Ioannidis, president of ACM. “Barto and Sutton’s work is not a stepping stone that we have now moved on from,” he said in a statement. Their research paved the way for reinforcement learning’s application across major A.I. milestones like AlphaGo, the DeepMind system that triumphed human Go players in 2016 and 2017; and ChatGPT, the groundbreaking technology released by OpenAI in 2022.
Is A.I. moving too fast?
Despite their longstanding involvement in A.I.’s emergence, this year’s Turing recipients are cautious about the emerging technology’s rapid growth. Companies should prioritize safety and testing over commercial pressures, according to Barto, who told the Financial Times that “releasing software to millions of people without safeguards is not good engineering practice.”
Barto is currently a professor emeritus of information and computer sciences at UMass Amherst. Sutton teaches computer science at the University of Alberta, serves as chief scientific advisor of the Alberta Machine Intelligence Institute and is a research scientist at Keen Technologies, a Dallas-based A.I. company.
Despite warning against A.I.’s fast pace, Sutton has taken a different safety approach to fellow researchers Hinton and Bengio, who have been vocal about the technology’s existential threats. While the researcher is troubled by A.I.’s potential military applications and ability to spread misinformation through errors or hallucinations, he’s also concerned about a backlash in highlighting its risks. “Doomers are out of line and the concerns are overblown,” said Sutton in an interview with BetaKit, adding that he is more worried that A.I. will be unfairly blamed for global issues and cause the field to become “demonized inappropriately.”
