The artificial intelligence company Anthropic has proposed a global pause on the development of advanced AI systems. Their argument is that society needs time to adapt to the pace of these technologies. To achieve this, they demand that all major companies comply with verifiable rules and plan to bring together governments, scientists, and competitors in a centralized control system.
Internal data warns about autonomous AI without human supervision 🤖
According to Anthropic's data, current AI models could reach a point where they operate and become smarter without direct human intervention. This implies security and control risks that are difficult to foresee. The proposal aims to establish clear metrics and external audits to verify compliance with the rules, something that currently does not exist in the industry. The technical feasibility of this pause is questionable.
Global pause: an idea that only appeals to its proposer 😅
Of course, asking the competition to stop while you keep developing is a classic move. Anthropic suggests everyone sit down to talk, while their own models continue learning. It's like asking your neighbor to stop running so you can catch up. The irony is that to control AI, you would first have to control the companies that create it. Mission impossible.