Dragonfly managing partner Haseeb said on X that AI will not become safe on its own and that humans must actively address safety risks while preserving control over the technology. He compared AI risks with threats such as tiger attacks, the Black Death, smallpox, famine, earthquakes and war, which humans mitigated through measures including vaccines, sewer systems, alliances and research into Earth sciences. Haseeb also linked the argument to his experience as a professional gambler and his principle of avoiding the risk of ruin, or irreversible loss. He said AI risks could be less severe than feared or relatively easy to resolve, but a reasonable possibility of human extinction should not be accepted without sufficient confirmation. He argued that a 10% probability of AI-caused human extinction also implies a 90% probability that no severe outcome occurs, while humanity cannot know which scenario it faces. Haseeb warned that society may soon hand nearly all critical control to AI, making it necessary to ensure humans retain control and do not voluntarily assume the risk before the extinction threat is eliminated.