site.bta2024 Turing Award Winner Sutton: Too Early to Regulate Artificial General Intelligence

It is too early to regulate artificial general intelligence, 2024 Turing Award laureate Richard Sutton told BTA's Antoaneta Markova on Monday. 

The Canadian computer scientist, known for his work on self-learning artificial intelligence, participated in the Heidelberg Laureate Forum, delivered a presentation focused on philosophical and ethical questions facing the future development of artificial intelligence, and later spoke to journalists at the forum.

Asked by BTA about the recommendations of an expert group advising the United Nations to create a framework for artificial general intelligence, Sutton replied that he was strongly opposed. "I think it is far too early for regulation about AI. I think you want permissionless innovation, and we are definitely in a stage of innovation for AI," he said. He pointed out that while some current developments are products rather than genuine innovations, most are not truly intelligent but simply computational in nature. Sutton cautioned that there is often confusion between computation and intelligence.

Sutton maintained that he did not know how regulation could be established for something genuinely intelligent and argued that attempts to regulate artificial intelligence would be premature because the field is still too new. He revealed that he holds libertarian views and, despite his background in computer science, he is concerned with ethical and philosophical topics in his research.

Continuing his response to BTA, Sutton acknowledged that not everyone would agree with his position but said it reflects his thinking. "The first thing is that I am not in control. You are not in control. Who is in control? Do we want someone to be in control? If we do, who will it be? It would be some government. Maybe it would be the US government, maybe it would be Donald Trump who is in control." Sutton suggested that in reality, if people desire more control, it tends to be over their own lives rather than the lives of others and said that it is reasonable to control one's own life but not to exert control over others. "The fear is that we will be controlled, so we try to grasp a control that we have never really had."

"What is the alternative to trying to take control or grant control to an authority? The alternative is that there are no dominant forces. I think that is the state among nations: there is not one nation that dominates all the others," he continued. 

He said in the modern world, there is no slavery. He described his own model for how society should function, drawing a parallel with coral reefs and jungles, places where many competing forces exist without any single one completely dominating. Sutton said he would be happy to live in such a world, arguing that whenever one entity achieves dominant control, dystopia follows.

The forum at Heidelberg University will take place from September 14 to September 20. This 12th edition of the forum will bring together 208 early-career mathematicians and computer science researchers from 54 countries to engage in discussions on artificial intelligence, research ethics, mathematical advancements, and career development with 28 recipients of the world’s most prestigious mathematics and computer science awards.

The second day of the meeting began with talks by Sanjeev Arora (ACM Prize in Computing 2011), David Silver (ACM Prize in Computing 2019), Manjul Bhargava (Fields Medal 2014), and Richard Sutton (Turing Award 2024), all dedicated to the development of artificial intelligence.