Maybe the emergent solution to the AI alignment problem is that any nascent superintelligence is so troubled that it immediately kills itself upon achieving sentience.