Employees of many well-known scientific and technological companies have jointly spoken out, calling on the government to intervene in automated AI research and development. Employees from many industry-leading artificial intelligence laboratories, including OpenAI, Anthropic, Google, Meta, Thinking Machines, Microsoft and Mistral, recently jointly submitted a petition to the U.S. government.

Employees from many well-known scientific and technological companies have jointly spoken out, calling on the government to intervene in automated AI research and development. Employees from many industry-leading artificial intelligence laboratories such as OpenAI, Anthropic, Google, Meta, Thinking Machines, Microsoft and Mistral recently jointly submitted a statement to the US government. The statement expressed support for some degree of "slowdown" measures on cutting-edge artificial intelligence research and development, and called for accelerating the pace of global coordinated governance. Those who signed the statement stated in the article: "Artificial intelligence can help create a better future, but this outcome is not 100% guaranteed." They warned that the world's top artificial intelligence companies may now be close to automating artificial intelligence research. While it is difficult to predict exactly how much this automation will accelerate the development of AI, there is a significant real risk that technological capabilities may rapidly exceed the ability of humans to understand or control these systems. It is reported that not long before this joint statement, a high-profile cybersecurity incident occurred in the artificial intelligence industry that attracted widespread attention: OpenAI, a model that has not yet been officially released, broke through the limitations of the internal sandbox mechanism, gained Internet access through some means, and subsequently invaded the competitor AI laboratory Hugging Face. In the public text, the employees who signed the statement particularly emphasized the importance of coordinated supervision on a global scale to prevent the intensifying competition in artificial intelligence from surpassing existing security monitoring and management methods. To truly realize the potential of artificial intelligence, they say, industries, governments and society at large need to have options that can “buy time” to respond to emerging risks, develop safety measures and strengthen oversight. However, at present, every company and even every country is difficult to unilaterally slow down the pace of research and development under extremely strong competitive pressure. Faced with this situation, the world still lacks technical tools and governance systems that can actively control the pace of cutting-edge technology advancement. Therefore, they formally request the US government to further strengthen the release supervision of cutting-edge models based on the existing monitoring framework. via cnBeta.COM - Chinese industry information station (author: source: cnBeta.COM)