🕒 Created · Updated

Sam Altman proposes industrywide slowdown in artificial intelligence development to address growing safety concerns

OpenAI CEO Sam Altman told employees that the firm is considering slowing down the development of cutting-edge artificial intelligence. He expressed hope that other AI labs would join the effort, though he acknowledged that some may not. The move comes as concerns grow regarding the potential for AI systems to autonomously design and develop other AI systems, which could lead to a user losing control of the resulting systems. To address these risks, OpenAI has already delayed the development and launch of its Astra model to test safeguards against cyber misuse and unauthorized actions. The company determined that Astra was the first model to meet a "Critical" cybersecurity threshold, meaning it could find and exploit security flaws across well-protected systems without human guidance. The administration announced that it is seeking to establish international coordination on future AI development. Meanwhile, researchers have suggested that developers should deliberately pace the frontier of automated AI development to allow societal structures and alignment research to keep up with the technology.

Sources

Paywall and unreadable sources