World leaders are being urged to take a more serious approach to the rapidly evolving field of artificial intelligence, as experts warn that not enough progress has been made since the last AI Safety Summit six months ago.
A group of 25 leading AI scientists, including Turing award winners and Nobel laureates, published a consensus paper calling for stronger action from governments in regulating AI technology. They stress the need for rapid-response institutions, increased funding, and more rigorous risk assessments with enforceable consequences.
According to Professor Stuart Russell from the University of California at Berkeley, it is time for strict regulation by governments rather than voluntary codes of conduct set by the industry. He emphasized the importance of understanding how to make advanced AI systems safe before further increasing their capabilities.
The experts highlighted both the potential benefits and risks associated with AI, such as boosting prosperity and scientific research on one hand, and enabling disinformation, job disruption, and inequality on the other. They also expressed concerns about the uncertainty surrounding the future of AI technology and the lack of consensus among experts on various critical issues.
The urgent call for action comes as tech giants like OpenAI, Google, Microsoft, and Apple continue to unveil new AI-powered tools and products, fueling the pace of innovation in the sector. As world leaders gather at the AI Seoul Summit, discussions will revolve around the interim report on AI safety, which aims to guide policymakers, industry experts, and researchers in addressing the complex challenges posed by AI.
Prime Minister Rishi Sunak will co-host the virtual summit with South Korean president Yoon Suk Yeol, emphasizing the responsibility governments have in managing the risks posed by AI technology. The global consensus among leading AI scientists indicates a pressing need for concrete commitments and stricter regulation to ensure the safe development and deployment of AI systems.
