I was pretty impressed by AI Risk is like Terminator; Stop Saying it's Not and the Followup, although both the author and I agree that it might potentially not be the best strategy for AI communication.

However, I did not see the post, or any of the comments, mentioning the fact that

Many technologies started out as science fiction before being invented.

It seems to me like a few people have thought about this, but never went out and told as many people as possible that this is one of the first things you can mention when explaining AGI to people. It seems obvious in retrospect (if you pay people to spend years writing and thinking about plausible future technology, they'll often land some solid hits, even if it's in the 1800s), but nobody mentioned it in the post and I've never heard it before. It's pretty clear that AI safety fieldbuilding is bottlenecked by the absurdity heuristic, everyone who ever once tried to talk about AGI with someone understands this personally.

It's probably not mentioned in The Precipice, and it doesn't seem to be mentioned in Superintelligence or WWOTF (note: this is from looking for "science fiction" at the index, not a search in the app, as I only have physical copies). The closest thing I could find anywhere was The Track Record of Futurists Seems Fine which was more of a forecasting kind of thing that evaluated predictions of various science fiction authors. There is only the disdain for science fiction that was first criticized in AI Risk is Like Terminator; Stop Saying It's Not. 

I think it's a good idea to put a list of technologies that started out as science fiction, and were then subsequently invented. It might even be valuable for AI safety people to just straight-up memorize the list, because we truly do live in a world where AGI strongly resembles science fiction, and we also live in a world where most people spend a few hours a day exposed to fiction.

There is a wikipedia list of technologies that started out as science fiction and you can send that link to people.

Technologies that started out as science fiction long before they became real:

  1. Nuclear bombs (1914, The World Set Free)
    1. This is more important than the rest of the list combined, and I recommend memorizing "The World Set Free" + "1914" and also that the book was read by Leo Szilard who played a major role in triggering the Manhattan Project.
    2. It's probably best to only memorize the details for nuclear weapons, and then just the inventions, because reciting a long list will probably come off as odd.
  2. The Internet (1898, From the "London Times" of 1904)
  3. Computer Screen (1878, a fake news article by Louis Figuier)
  4. Space Travel (1657, Comical History of the States and Empires of the Moon)
  5. Video Calls (1889, In the Year 2889)
  6. Aircraft (Various kite-flying enthusiasts in ancient China, and then Leonardo Da Vinci in the 1400s)
  7. Computers (1726, Gulliver's Travels, the first computer programs were designed by Ada Lovelace and Charles Babbage in the 1840s and the first electric computers were built in the 1930s)


New Comment
2 comments, sorted by Click to highlight new comments since: Today at 4:11 AM

I'm not sure I understand the case for this being so urgently important. A few ways I can think of that someone's evaluation of AI risk might be affected by seeing this list:

  1. They reason that science fiction does not reflect reality, therefore if something appears in science fiction, it will not happen in real life, and this list provides lots of counterexamples to that argument
  2. Their absurdity heuristic, operating at the gut level, assigns extra absurdity to something they've seen in science fiction, so seeing this list will train their gut to see sci-fi stuff as a real possibility
  3. This list makes them think that the base rate for sci-fi tech becoming real is high, so they take Terminator as evidence that AGI doom is likely
  4. They think AGI worries are absurd for other reasons and that the only reason anyone takes AGI seriously is because they saw it in science fiction. They also think that the obvious silliness of Terminator makes belief in such scenarios less reasonable. This list reminds them that there's a lot of silly sci-fi that convinced some people about future tech/threats, which nonetheless became real, at least in some limited sense.

My guess is that some people are doing something similar to 1 and 2, but they're mostly not people I talk to. I'm not all that optimistic about such lists working for 1 or 2, but it seems worth trying. Even if 3 works, I do not think we should encourage it, because it is terrible reasoning. I think 4 is kind of common, and sharing this list might help. But I think the more important issue there is the "they think AGI worries are absurd for other reasons" part.

The reason I don't find this list very compelling is that I don't think you can look at just a list of technologies that were mentioned in sci-fi in some way before they were real and learn very much about reality. The details are important and looking at the directed energy weapons in The War of the Worlds and comparing it to actual lasers doesn't feel to me like an update toward "the future will be like Terminator".

(To be clear, I do think it's worthwhile to see how predictions about future technology have fared, and I think sci-fi should be part of that)

Nuclear submarines (1870, Twenty Thousand Leagues Under the Sea)

Time travel (1895, The Time Machine)

New to LessWrong?