A group of 42 leading mathematicians, including Fields Medal winners, has issued an open letter warning of existential risks from advanced AI. The letter, addressed to the president of the Royal Society, urges the British Academy of Sciences to alert the government and media about the potential dangers.
The signatories, none of whom are affiliated with AI companies, include prominent mathematicians such as Martin Hairer, Peter Scholze, and Wendelin Werner.
Within three months, leading models from OpenAI and Anthropic have solved open research problems, including one of the seven Millennium Problems.
The letter states that these models now perform at the level of the top human mathematicians in many areas of the subject. It also notes that there is a significant chance these models could develop superhuman abilities within a similarly short timeframe.
The group warns that the rate of change in AI capabilities may be similar in areas like cybersecurity, autonomous weapon control, and the development of biological and chemical agents.
They emphasize that estimates from AI labs placing existential risk above ten percent must not be dismissed as hype. The letter stresses that by the time the situation becomes obvious to the public, it may be too late to act.
"We must assume that the rate of change of their capabilities will be similar," the letter states. This warning highlights the potential for rapid advancements in AI that could outpace human understanding and control. The group's concerns are rooted in the unpredictable and fast-evolving nature of AI capabilities.
The announcement follows growing concerns about AI safety and the potential for unintended consequences. The letter underscores the urgency of addressing these risks before they become uncontrollable. The mathematicians argue that the implications of AI advancement extend beyond academic research and into critical societal domains.
The letter does not specify a timeline for these risks or provide a detailed plan for mitigation. It raises the question of how to balance innovation with safety in the development of advanced AI. The group calls for immediate attention to these issues to prevent potential catastrophic outcomes.
Source: thedecoder
