Giving agents bounded autonomy - InfoWorld

Discussion of how to give AI systems limited decision-making power while keeping humans in control. This touches on important questions about AI safety in automated systems that people increasingly interact with.