work with Windows. :grin:
It may not be terribly relevant, but one story I love to tell in these sorts of conversations has to do with a time that Douglas Hoftadter was asked to put an Army project to the Turing test. He sat down at the terminal, and, fairly quickly, "unmasked" the computer.
Except it wasn't a computer. It was a couple of students at a teletype in another room. The transcript, with commentary from Dr. Hoftadter, can be found in The Mind's I, or in Gödel, Escher Bach, or in Fluid Concepts & Creative Analogies--I don't recall which one, and my Goggle-fu isn't being helpful on it, I'm afraid--and makes for a fascinating read. By the end of it, Dr. Hofstadter had figured out he was being had, but only after the students on the teletype hit him with very blatantly odd answers in the context of the Turing test.
The point of which digression, I guess, is that we must be most leary of our assumptions, if we are seeking to test an intelligence. Assume it is just a machine, and it turns out to be two grad students; assume the taker of the test is a moderately handicapped idiot, and it turns out he has an IQ of 160 but simply doesn't speak, read, or write the English in which the test's instructions are written. How often have you seen an unsuspecting user taken in by a chat-bot?
Look at what the intelligence does, though, and it reveals itself. The two students fooled the doctor until they began trying to get "cute" with their answers. The genius who didn't speak English still figured out enough of the IQ test to look like a somewhat deficient human mentality on a test in a language he could not speak, read, or write.
Most computers and applications today can't qualify as intelligent by any measure, but they're not written to be such. The ones that are, though, are stunning in their abilities, and, in the performance of the tasks for which they were written, often (unsurprisingly) vastly exceed human capabilities. The field is still very underpopulated in strong AIs--the self is very elusive, it turns out--but the weak AIs that dominate it can be very surprising in what they can do.
When you say "it's taken as a given that none of the posts are originated from within the forum," you reveal your assumptions. You are already starting from the assumption that everything you read, here, is posted by another human being who thinks more or less as you do. But, what if it's not? What if, somewhere among all the UFies in the world, is one that "lives" by electrical impulses through an array of processors, instead of having a brain, central nervous system, glands, and a bunch of wierd chemicals floating in amongst all of that? If this electronic intelligence were self-aware, responsive to its environment, and capable of operating under conditions where it must cope with the unknown, would you be able to pick out which one it was? |
