AI superintelligence, viewed as humanity's pinnacle invention, could help solve many world-critical issues. However, its extensive power also carries potential dangers, which could even lead to humanity's disenfranchisement or extinction. As it seems distant now, experts predict it may be realized within the coming decade. To manage these risks, new governance institutions and the resolution of the superintelligence alignment problem are needed. The question at hand is how to ensure these AI systems, which could surpass humans in intelligence, follow human intention. Currently, there are no viable solutions to control or direct potentially superintelligent AI, therefore stopping it from going rogue. Existing AI alignment methods, like reinforcement learning from human feedback, hinge on humans' ability to supervise AI. However, it's improbable that humans can consistently supervise AI systems that are significantly smarter than us, and hence our current alignment methods will not scale to superintelligence. Therefore, new scientific and technical breakthroughs are required.