The intrinsic value of human perspectives is at stake as superintelligent AI poses inherent risks such as misaligned objectives and an intelligence explosion that could end human dominance, making it imperative to solve control problems.