The editorial board argues that humanity successfully restrained nuclear technology after initial fears, offering a blueprint for governing artificial intelligence. Just as President Harry Truman faced the dual potential of nuclear power for destruction or flourishing in the late 1940s, policymakers today confront similar dilemmas with AI. Prominent voices warn of existential risks, including AI agents acting autonomously and the critical issue of "alignment" — ensuring AI prioritizes human values. Executives from leading firms, including Anthropic's Dario Amodei, have called for development slowdowns. The piece emphasizes that government regulation, not self-policing by tech companies, is the proven model for managing powerful technologies. History shows sustained, deliberate effort can mitigate dangers, and the same cooperation is necessary for AI safety today.

The overriding principle is that companies building AI cannot be left to police themselves. Just as the military developed atomic energy initially, the private sector now leads AI advances, but regulation must follow. The editorial warns against allowing a small circle of profit-driven actors to dictate rules. Lessons from the nuclear era demonstrate that society can restrain dangerous innovations when it decides to act, provided there is sustained political will and international cooperation to ensure the technology serves humanity's future rather than its demise.