AI: The Species Metaphor
Let’s kick off my new blog with a post that broadly slices through two of my nerd interests, Artificial Intelligence, and Science Fiction. Though, these days of course, the borders between them are blurrier than ever.
I first came across the AI: Species Metaphor about ten years ago when it was briefly covered in Dan Carlin’s sadly now dormant ‘Common Sense’ Podcast. The episode in question, titled ‘Summoning the Demon’ follows the narrative of an essay by prominent AI researcher (and cautionary voice) Eliezer Yudkowsky.
In a nutshell, the AI Species Metaphor is an attempt to encapsulate what a ‘Super Intelligence’ might seem like once it soars past the limits of human comprehension.
This leap would occur during what has been termed an “Intelligence Explosion,” which refers to the moment an artificial intelligence rapidly becomes smarter than us. Most people tend to rationalise “smart intelligence” as the spectrum between “the village idiot” and Einstein, which is a prime example of anthropomorphic thinking. If you place these two on a map of all possible intelligences, the gap between them is tiny.
So how, then, might we attempt to understand (or fail to understand) the difference between human civilization and a newly developed (and self-improving) Super Intelligence? Yudkowsky poetically writes:
“There is finally the family of species metaphors, based on between-species differences of intelligence. The AI has magic - not in the sense of incantations and potions, but in the sense that a wolf cannot understand how a gun works, or what sort of effort goes into making a gun, or the nature of that human power that lets us invent guns.”
Yudkowsky then goes on to quote a fascinating snippet from 1993 essay by the late Vernor Vinge, the distinguished Computer Scientist, Sci-Fi author.
“Strong superhumanity would be more than cranking up the clock speed on a human equivalent mind. It’s hard to say precisely what strong superhumanity would be like, but the difference appears to be profound. Imagine running a dog mind at very high speed. Would a thousand years of doggy living add up to any human insight?”
So there you have it, try to imagine human experience and ingenuity from the perspective of a Canine or Canis mind, and you’ll perhaps catch a glimpse of the unfathomable insight that a god-like Super Intelligence might possess.
Quotes taken from Nick Bostrom and Milan M. Ćirković’s apocalyptic compendium, Global Catastrophic Risks ( Oxford University Press 2008).
