Andrew Barto and Richard Sutton won the 2024 Turing Award for pioneering reinforcement learning, a key AI technique. They warned against rushed AI deployment without safeguards and criticised commercial motives in AI. Sutton dismissed AGI hype, while both highlighted risks of funding cuts to research and the need for responsible AI development.
The Association for Computing Machinery (ACM) has awarded the 2024 Turing Award to Andrew Barto and Richard Sutton for their foundational contributions to reinforcement learning (RL), a key machine learning technique that has driven the modern artificial intelligence (AI) boom. The award, often referred to as the “Nobel Prize of Computing”, comes with a $1 million prize, sponsored by Google.
Barto, a professor emeritus at the University of Massachusetts, and Sutton, a professor at the University of Alberta and former research scientist at DeepMind, are credited with introducing reinforcement learning principles, developing their mathematical foundations, and designing key algorithms. Their work has significantly influenced AI developments, leading to innovations such as Google DeepMind’s AlphaGo and OpenAI’s ChatGPT.
The foundations of reinforcement learning
Reinforcement learning, a branch of machine learning, is inspired by psychology and the way humans learn through trial and error. It involves training AI systems to make optimised decisions by rewarding desired behaviours. Barto, a professor emeritus at the University of Massachusetts, and Sutton, a professor at the University of Alberta and former research scientist at DeepMind, developed the mathematical foundations and key algorithms for RL in the 1980s.
Their work has since become a central pillar of the AI boom, driving billions of dollars in investment and enabling breakthroughs in fields ranging from gaming to natural language processing.
Their seminal textbook, Reinforcement Learning: An Introduction, has been cited over 75,000 times and remains a foundational resource for researchers. Jeff Dean, Senior Vice-President at Google, which sponsors the Turing Award, praised their work, stating, “The tools they developed remain a central pillar of the AI boom and have rendered major advances, attracted legions of young researchers, and driven billions of dollars in investments. Reinforcement learning’s impact will continue well into the future.”
Concerns over unsafe AI deployment
Despite their optimism about the potential of RL, Barto and Sutton have expressed serious concerns about the current trajectory of AI development. They argue that many companies are rushing to release powerful but untested AI models, prioritising speed over safety.
“Releasing software to millions of people without safeguards is not good engineering practice,” Barto told The Financial Times. He likened the approach to “building a bridge and testing it by having people use it.” He added, “Engineering practice has evolved to try to mitigate the negative consequences of technology, and I don’t see that being practised by the companies that are developing.”
Sutton echoed these concerns, criticising the industry’s focus on business incentives rather than advancing AI research. “The idea of having huge data centres and then charging a certain amount to use the software is motivating things, and that is not the motive that I would subscribe to,” he said.
The debate over artificial general intelligence
The scientists also weighed in on the debate surrounding artificial general intelligence (AGI), a term used to describe AI systems that match or surpass human cognitive abilities across a wide range of tasks. Sutton dismissed the term as “hype,” arguing that the concept of AGI is often misunderstood.
“AGI is a weird term because there’s always been AI and people trying to understand intelligence,” he said. He believes that achieving systems more intelligent than humans will require a deeper understanding of the human mind, rather than simply scaling up existing technologies.
OpenAI, one of the leading AI companies, has defended its focus on AGI, stating that its mission is to ensure that such systems benefit humanity. However, Sutton remains sceptical, arguing that the current emphasis on large language models (LLMs) like ChatGPT may not be the right path to achieving true intelligence. “I don’t think that’s the direction that’s going to lead to full intelligence,” he said.
Yoshua Bengio and Geoffrey Hinton, who were awarded the Turing Award in 2018 for their work on deep learning, have previously voiced strong concerns over AI risks. A 2023 statement signed by AI leaders, including OpenAI CEO Sam Altman, warned that “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
The Alberta Plan: A roadmap to full intelligence
Richard Sutton and his colleagues at the University of Alberta have introduced The Alberta Plan, a 12-stage roadmap aimed at achieving what they term “full intelligence.” This structured approach focuses on developing AI systems capable of fully understanding and interacting with their environment.
The plan outlines key steps towards building advanced AI models that mirror human cognitive abilities. Sutton estimates a 25% probability of reaching human-like intelligence by 2030, with the likelihood increasing to 50% by 2040, as computing power becomes more affordable and efficient. The initiative represents a systematic effort to push AI research towards achieving true general intelligence.
AI safety and ethical concerns
While Sutton is critical of the term “AI safety,” he acknowledges the risks associated with deploying AI systems that are prone to errors and hallucinations. “I’m disappointed that my fellow researchers are playing into the way their field is possibly going to be demonised inappropriately,” he said. He believes that the real issue lies in how people use AI tools, rather than the technology itself.
“That’s not really a problem with the technology—it’s a problem with people being gullible,” he argued. Sutton also expressed concerns about the potential for AI to be used in military applications, stating, “We shouldn’t be eager to make force-projecting AI.”
The AI sector is experiencing record-breaking investments, with Big Tech companies expected to spend over $320 billion on AI development this year. OpenAI alone is raising $40 billion in new funding at a $260 billion valuation.
The future of AI and reinforcement learning
Despite their concerns, both Barto and Sutton remain optimistic about the potential of AI and reinforcement learning to bring positive outcomes to the world. “We have the potential to become less greedy and selfish and more aware of what’s going on in others… there are many things wrong in the world, but too much intelligence is not one of them,” Sutton said.
As the field of AI continues to evolve, the work of pioneers like Barto and Sutton will remain foundational. Their contributions have not only advanced the technology but also sparked important conversations about its ethical and safe deployment.
Pallavi Singal is the Vice President of Content at ztudium, where she leads innovative content strategies and oversees the development of high-impact editorial initiatives. With a strong background in digital media and a passion for storytelling, Pallavi plays a pivotal role in scaling the content operations for ztudium’s platforms, including Businessabc, Citiesabc, and IntelligentHQ, Wisdomia.ai, MStores, and many others. Her expertise spans content creation, SEO, and digital marketing, driving engagement and growth across multiple channels. Pallavi’s work is characterised by a keen insight into emerging trends in business, technologies like AI, blockchain, metaverse and others, and society, making her a trusted voice in the industry.
