Self-improving AI and the risk of losing human control
Posts citing DRM News’ YouTube channel quote Sam Altman warning the UN Security Council that AI progress could outpace human oversight, and saying OpenAI has slowed development before and will again.
TLDR
Posts citing DRM News’ YouTube channel quote Sam Altman at the UN Security Council warning that humans could lose control if AI advances too quickly for people to follow or intervene. He calls for “extreme care” as systems approach recursive self-improvement—the ability to improve themselves and future versions—and says automating AI development could rapidly accelerate progress. The posts also quote him saying OpenAI has “unilaterally slowed down in the past” and will do so again.
