Chimpanzees and primates in general:

A listening policymaker or tech executive might think:

  • I saw something about this on TV a while ago. The chimpanzee wins the fight every time.
  • A single human can't wipe out any primate species. There were all these social and civilization structures, and the chimpanzees were largely an externality to the market anyway. Are we talking about a new dominant race of AIs or something? Like Planet of the Apes? They're tools, you know.


Ants and bees:

A listening policymaker or tech executive might think:

  • Well, 99% of anthills never get noticed by people at all.
  • Insects are actually much more successful than humans, as far as biomass goes. Did you know that mammals are actually one thousand times more likely to go extinct first?
  • Ants and bees are much more coordinated than humans. If you had the experience that I did, you'd know how much better and wealthier the world would be if we could keep our workplace offices as simple and efficient as ant and bee colony organization.
  • Long before bees actually went extinct, "saving the bees" quickly became the most popular thought that anyone has about any insect.


Dogs, cats, etc:

A listening policymaker or tech executive might think:

  • Ew
  • Who is this person
  • Why are they in my office
  • Alternatively: You realize that my business/policy domain revolves around AI, right? Are you trying to start something? Because that's a pretty rude, bizarre, and provably incorrect way to describe today's AI industry.


Livestock and lab rats:

A listening policymaker or tech executive might think:

  • Most people who own livestock choose to consider them as pets and treat them humanely.
  • A single oil tanker contains more chemical energy than all the people in the entire state/province.
  • Lab rats and experimentation in general tends to save many more lives than it harms, since the results scale to large numbers of people.


Other bad or potentially risky intelligence comparison examples include:

  • Dodos
  • Hippos
  • Peacocks
  • Animals that are so much simpler than humans that extrapolating that difference to the difference between humans and AGI causes a more intense absurdity heuristic than it needs to be; for example, why invoke a broad, vague sort of feeling of something that's as beyond us as we are to ants, when you can just keep it simple e.g. a cat doesn't know that it's going to die, but humans know that that cat will die? You never really know who you're dealing with and what's cognitively available to them at that moment, and you really don't want them to lean on the concept of divine beings (any more than the risk can be avoided in the first place).

Generally, explaining AGI is a difficult task, but comparing human intelligence to animal intelligence is increasingly vital to keeping it quickly and efficiently understandable. Especially under ordinary conditions, when you are highly uncertain what the background of a person is and what kinds of simpler concepts they're likely to lean on as a crutch during the critical first 30 seconds of explaining the concept of AGI.


New Comment