Artificial intelligence pioneer Geoffrey Hinton has called on governments worldwide to collaborate in training AI systems to avoid potential harm to humanity.
At the World Artificial Intelligence Conference (WAIC) in Shanghai, Hinton emphasized the need for international cooperation on non-hostile AI development, likening it to Cold War-era nuclear non-proliferation efforts. He warned that unless nations work together, the rapidly advancing technology could surpass human intelligence and pose significant risks.
Speaking at the conference, Hinton described AI development as akin to raising a tiger cub, cautioning that if nations fail to train it properly, it could become a threat. He pointed out that while offensive AI use, such as cyberattacks and lethal autonomous weapons, is unlikely to be widely cooperative, there is potential for nations to form a network of institutions to guide AI’s development towards non-hostile outcomes.
Hinton further highlighted the importance of ensuring AI systems do not seek to eliminate humans, noting that simply turning off AI once it surpasses human intelligence would not be an option. He also mentioned that while rational nations may find cooperation relatively easy, the current U.S. administration could present challenges to such efforts.
In parallel, Beijing has announced its intention to establish an organization to coordinate international AI cooperation, as stated by Chinese Premier Li Qiang during the WAIC. The Chinese leadership has underscored the necessity of forming a global AI governance framework to ensure responsible development and use of the technology.
Hinton’s remarks have sparked discussions on the ethical and strategic implications of AI development. As nations continue to invest heavily in AI research and application, the call for collaboration becomes increasingly urgent, with the potential to shape the future of technology and its impact on humanity.