Standard Futuristic Dystopia #43-C: Artificial superintelligence takes over humanity by force and enslaves us into forced labor or worse.
This is a time-honored scenario, reworked into stories from R.U.R. to Wall-E. It's also slightly less likely than Standard Futuristic Dystopia #683-A: "Pyrokinetic produce incinerates the world's supermarkets."
But what about reality? When, in history, has intelligence done more harm than mindlessness? Countries have been cowed into obeisance by the strong, the merciless, the charismatic and the rich, but rarely do you see a résumé that includes both important contributions to astrophysics and the phrase "ruthless overlord to millions of dispirited serfs."
To be precise, the best and brainiest have certainly contributed to their share of revolutions, but they usually end up beheaded in the first post-revolutionary purge, proving that book smarts aren't as important as guillotine smarts.
This doesn't mean that we should just go ahead and enbrainify our machinery with no concern for consequences, though.
Behold the Apocalyptic Valley
My theory is that the AI curve looks something like this:
That dip you see there is what I call the Apocalyptic Valley.
Up to a certain point, the more intelligent your software, the better. For instance, if my spam blocker ever figures out that I've never done business with anyone who addresses me as "Dearest Gentleman or Madam," I'll be much happier.
Similarly, if a superintelligent computer, one smarter than any human ever, can be talked into ruling the Earth, we'll all be better off. (I will probably spend my time creating hyper-Sudoku puzzles only it can solve.)
Somewhere in the middle, computers will advance to the point where everyone will have, on their desk or in their pocket, the equivalent of an obedient idiot. And that's when we'll be in trouble.
A single supersmart person is no threat to anyone except a rival Warcraft guild. A fairly intelligent person with an army of obedient idiots is the most dangerous thing on the planet.
There are lots of ways that various idiot-toting would-be tyrants could destroy civilization, but here's the one that comes most readily to mind: spam.
A fairly intelligent person with an army of obedient idiots is the most dangerous thing on the planet.Spam is the sun in August, and the internet is a redhead slathered in SPF 50. Spam is the ocean, and the internet is Denmark. Spam is the endless vacuum of space, and the internet is a Russian spacecraft with a dog in it.
Only two things make online communication viable: the slowness of humans and the dumbness of computers. Humans can't spam more than one site every minute or so, and computers can't read fuzzy images of made-up words like "CRePSclE."
Once that gap closes -- once someone invents AI that's as fast as a computer and as smart as the dumbest literate person on the planet -- the walls will tumble and we'll be up to our e-mail addresses in advertisements for things nobody needs. Legitimate businesses will need to follow suit just to tread water, and civilization will collapse under the weight of bogus antivirus software and herbal Viagra.
Some say the world will end in fire, some say in ice. I say it will end in spam.
Image: Lore Sjöberg, incorporating an image by Alan Levine/Flickr
- - -
Born helpless, nude and unable to provide for himself, Lore Sjöberg eventually overcame these handicaps to become a posturing, obnoxious-apocalypse-postulating prognosticator.
See Also:- Alt Text: The Sorry De-Evolution of Spam
