Everyone is talking, or debating, about so-called “Artificial Intelligence.”
“AI” is smarter than humans; it will take over the whole world as we know it; humans will become useless, etc., etc. …
My admittedly amateur view is that whatever this is, it should be called “Artificial Calculation” rather than “Artificial Intelligence.”
My debating opponents argue that AI is capable of “thinking” for itself (programming itself) and coming up with novel solutions to all sorts of problems.
I reply that it’s just a form of machine calculation made possible by the human programmers who feed the machine its program(s). The machine can only create novel outcomes made possible by the program. There is no “intelligence” there, independent of what the programmers have made possible.
One very enthusiastic friend argues that within a decade, a computer will be able to compose an entire play every bit as great as Shakespeare’s King Lear, while another party to this debate, says “No way!”
Part of my argument, besides the claim that computers are just slaves to their programmers, and are just calculating, not “thinking,” is that a computer is just an obedient machine, a house with nobody home.
Another hot topic has to do with AI and what most people call “Ethics”. How can a machine distinguish right from wrong. How will a self-driving car, for example, which may be in a situation where it cannot avoid hitting something, choose between hitting the child, or the pregnant woman passing in front of it?
On that topic, here is a particularly good piece Joseph Brean wrote for The National Post, (June 1).