Power Utilitarianism Continued (Pt.3) -- Why We 'Should' Release Superintelligent AI into the Wild