OpenAI Proposes Global Standards for AI Alignment and RSI Safety

OpenAI wants global standards for frontier AI safety, warning that careless recursive self-improvement could cost humans control over AI development.