‘Scorpions in a Bottle’: What the U.N. Can Do About AI Safety

This article discusses the United Nations General Assembly's (UNGA) high-level week, where leaders and diplomats are focusing on the risks posed by artificial intelligence (AI) and the United Nations' (U.N.) role in addressing this issue. Amandeep Singh Gill, the U.N. undersecretary-general and special envoy for digital and emerging technologies, has been working on AI governance and has hosted a "Digital Cooperation Day" to discuss collaboration on AI opportunities and risk management.

Gill emphasized that the U.N. can serve as an inclusive platform for AI-related conversations, grounded in scientific research and best practices. The panel can provide information, best practices, and insights for audits and evaluations, allowing countries to learn from each other's experiences. While the U.N. can play a crucial role in facilitating these discussions, the involvement of major AI developers like the United States and China is essential to set the tone for the global community.

Gill highlighted that the U.S. and China possess 90% of the compute power and most of the investments in AI, making them influential in shaping the future of AI technology. He stressed that the U.N. needs to engage these countries and encourage transparency, information exchange, and the alignment of assessments of AI capabilities.

However, Gill also emphasized that the global community should not wait for the U.S. and China to take action, as many other member states are actively investing in AI and are concerned about safety and the opportunities that AI presents.