chain-of-thought monitoring

OpenAI Says AI Scaling Needs Stronger Safety Bars

OpenAI Chief Scientist Jakub Pachocki argues that rapid AI progress is making alignment and monitoring the central limits on responsible development.