How Long Until AI Doesn't Need Humans? (asteriskmag.com)

In a recent discussion on the future of artificial intelligence, Ajeya Cotra from METR and Timothy B. Lee from Understanding AI explored the contentious timeline for achieving “self-sufficient AI” — systems capable of maintaining and reproducing themselves without human intervention. Cotra believes this milestone could be reached within a decade, as advancements in AI could eventually enable systems to operate factories, mines, and other critical infrastructures autonomously. Conversely, Lee argues that significant physical and technical barriers mean we may not achieve true self-sufficiency for 50 years or more, if at all. He highlights that, while humanoid robots are currently being produced, they lack the dexterity and efficiency comparable to humans, and the economic viability of scaling their use remains an unresolved challenge. This debate is significant for the AI/ML community as it raises critical questions about the future roles of AI and humans in the workforce. If AI systems do become self-sufficient, they could fundamentally alter societal dynamics, raising concerns about potential conflicts with human interests. The conversation also underscores the importance of distinguishing between theoretical possibilities of AI autonomy and the practical realities that could delay those advancements, emphasizing that the path to truly capable humanoid robots and self-sustaining AI includes not only cognitive improvements but also advancements in physical hardware and maintenance logistics.