I worked at OpenAI. Here are the guardrails we need now | Miles Brundage | The Guardian

... AI systems' capabilities, risks and safety mitigations in detail. Experts ... risks of unbridled AI competition are widely recognized. The US ...