The NYT wants the US government to create a regulatory system for AI that mirrors the one created for nuclear energy after World War II. Does that make sense, and can it work?
Yes, but it will be much harder, because the bomb was purely a government project, while AI labs belong to incredibly wealthy tech titans with powerful incentives to escape regulation. In addition, the US government had the trust of its citizens after the war; do you really trust Trump to regulate AI wisely? Finally, adding layers of regulation would probably slow innovation and would expose the regulators to criticism about falling behind the Chinese. That will be a political football if there ever was one.
Of course, falling behind in the race to exterminate mankind is not necessarily a bad thing.