Comment by Roman V. Yampolskiy

Advanced AI, especially if achieving what many refer to as Artificial General Intelligence (AGI), would not be limited by the same temporal and cognitive constraints as humans. Current literature on AI risk often focuses on control problems and rapid FOOM (fast onset of overwhelming mastery) to doom scenarios—where a system quickly self-improves and becomes uncontrollable causing existential risks. Such scenarios emphasize either swift, dramatic adversarial moves or near-instant catastrophic failures. While these remain plausible, they overlook the advantage that long-term strategizing confers upon an immortal digital entity, which can optimize outcomes over spans that humans can scarcely conceive as relevant planning horizons.