The U.S. must keep racing China on AI because stopping unilaterally, as with Cold War nuclear arms, cedes advantage to adversaries
Source: https://www.facebook.com/americanspectator/. "The AI Race Is a Race We Cannot Afford to Lose | The American Spectator | USA News and Politics." October 20, 2026. spectator.org
The Gist
The author says every powerful technology has both good and bad uses, and the U.S. has always had to move ahead anyway because its rivals do. Since China is racing for AI dominance, the U.S. can't stop without letting China pull ahead, just as the Soviets gained ground when America paused its nuclear buildup.
Conclusion
The United States must continue to aggressively develop and exploit AI/superintelligence for civilian and military purposes rather than slow or stop, because it is in a geopolitical race with China that it cannot afford to lose.
Premises
- Like all major technologies (fire, metal, rocketry, nuclear power), AI will bring both benefits and harms, so the existence of serious risks does not by itself justify abandoning it.
- In the nuclear era, the U.S. pursued the atomic and hydrogen bombs despite real catastrophic risks and internal opposition (e.g., Oppenheimer), and this was vindicated because the U.S. beat the Soviets to the H-bomb.
- U.S. policy decisions must be made in a specific geopolitical environment, and the current environment is a race with Communist China for AI/SI superiority.
- If the U.S. stops racing, China will gain ground; the 1960s U.S. unilateral halt in building nuclear missiles let the Soviets reach parity and then pursue superiority and first-strike capability.
- Harold Brown's maxim, 'When we build, they build. When we stop, they build,' shows that adversaries do not reciprocate American restraint.
Assumptions
- The AI situation is sufficiently analogous to the Cold War nuclear arms race for its lessons to transfer.
- China is pursuing AI/SI superiority and would not reciprocate U.S. restraint or negotiate limits.
- AI/SI superiority confers decisive geopolitical, economic, and military advantages comparable to nuclear superiority.
- The only meaningful alternative to racing is unilateral U.S. stopping, not regulated or coordinated development.
- The risks of AI (including loss of human control) are manageable or are outweighed by the risk of falling behind.
- Winning the race will produce a safer outcome for the U.S. than losing it.