For question related to the "AI control problem".
In artificial intelligence (AI) and philosophy, the AI control problem is the hypothetical puzzle of how to build a superintelligent agent that will aid its creators, and avoid inadvertently building a superintelligence that will harm its creators.