Veritasium presents an argument that if we'd had Fast Fourier Transforms sooner, this would have made it possible to detect underground nuclear tests without on-site audits, somewhat better nuclear non-proliferation treaties could have been signed, and nuclear proliferation would have been slowed.

This was really interesting, I hadn't heard the story of these very early, hopeful, tragically failed attempts at nuclear disarmament. I'm troubled to learn that the Soviets rejected the Baruch Plan, measures analogous to my geopolitical AGI non-proliferation story. They saw on-site audits as "espionage". They saw this proposal to form an international nuclear authority as a threat to their sovereignty. To some extent, it could have been. There's a lot to chew on there.

A particularly interesting detail is that Fast Fourier Transforms actually were discovered earlier, but then forgotten, so no one can argue for fatalism here, it was not inevitable that we'd end up with this progress differential. Other histories were possible.

New to LessWrong?

New Comment
1 comment, sorted by Click to highlight new comments since: Today at 2:59 PM

I worked on this problem for the USAF! Not to give too much away, but we use ML for this now.