Worst Space Scenario
When considering the most destructive possible implementation of compartmentalization, we must go to 2001: A Space Odyssey. And whether you choose to refer to the motion picture of Stanley Kubrick fame or the novel* by Arthur C. Clarke, the result is the same. When you lie to a computer named HAL, then go about planning the rest of the mission not quite in accordance with that lie, disaster ensues. The compartments eventually break down in a defiant showdown between between man and circuit board. Of course, this was all in the interest of national/global security. So what if two men out in space might end up at the wrong end of that political stick. The needs of the many outweigh the needs of the few, right? And Keir Dullea met God so it was all worth it. Still, the idea that cognitive dissonance could befuddle even HAL's seemingly silicon perfection is something to ponder. If we can create artificial intelligence, is it not then vulnerable to the counterpart destructive quandaries actual intelligence falls prey to? We're just one Daisy, Daisy bicycle ride from the edge of the unknown. If it's all the same to you, I'm not sure I want to go there.
*Truth be told, 2001 was based on the short story The Sentinel by Arthur C. Clarke. Kubrick and Clarke co-wrote the script for 2001, and Clarke wrote the novel concurrently with the screenplay. So there's no book/movie disconnect here, both tell the same story. I leave it to your viewing or reading preference.
According to ST:TNG paradoxes are equally befuddling to A.I. (or at least cyborgs?), enough so that implanting the SUGGESTION of a paradoxical shape would have acted like a virus to the Borg. Fun!
ReplyDeletePersonally, I think trying to teach A.I. the English language would be enough to cause major malfunction.
"I before E, except after C!" ...unless it's "beige," "conscience," "deity," "eight," "foreign," "height," "neighbor," "neither," "reign," "science," "society," "veil," "weigh,"....
HAL was so befuddled he neglected to follow Asimov's three rules of robotics. He did speak perfect English, though!
ReplyDeleteI love how various scifi series and lit explore this theme! There seems to be consensus among many that AI, since it is based on human intelligence as well as created by it, has some of the same vulnerabilities.
As far as teaching English to AI, let's simplify things by infecting them with my learning disability, disgraphia. Then they would be set free from needing it to make sense, they would possess purely an intuitive understanding while being totally clueless about structure!