Presuming it is feasible to develop and employ systems with human or super-human levels of cognition and awareness, we draw on psychology, neuroscience, ethics, law, political economy, and artificial intelligence to define superintelligent AIs that are anti-authoritarian, ethical, and altruistic. We delineate the principles, milestones, and research necessary for such development, emphasizing the importance of interdisciplinary collaboration and innovation.