Lately I've been thinking about the idea of planting human intelligence in a computer. If you had asked me about it ten years ago I would have said it'll never happen, but now I'm not so sure.
First, some kind of thinking is already in computers, almost invisibly. Here's an example. I type a lot of text into the browser, Chrome, and it's talking back to the Google servers and is constantly evaluating what I type.
Over time it gets better at showing stuff that needs correcting. It started with spelling errors. Now it understands proper names and it can highlight some bad grammar.
Examples:
Its sunny today. (Not flagged, but it should be, the correct word is a contraction, "it's" for "it is.")
Bill Gates got her hair done today. (Bill Gates is masculine, should be flagged it should be "his hair.")
I object on principal. (Flagged with a gray underline. Good!)
Marylin Monroe. (Red underline, name misspelled.)
Yesterday I eat some food. (Not flagged, but "eat" is the wrong tense.)
The dog to the moon flew. (Bad grammar, not flagged.)
So this is good. I like it when my computers make my work smarter, automatically without me having to do anything. But I'm not sure this is thinking. It's a lot like data processing, something computers have been doing since the beginning.
The tough question is can we make computers think like a human and if we can should we want to? Much of what humans act on is unconscious emotion, not thought, based on how things feel. Should we teach computers to do that? Is that what we had in mind when we talk about computers thinking?
Pop out and look at the big picture. What's been accomplished with human thought? Where did all this thinking lead us to? An over-populated unsustainable civilization. More thinking like that? I hope not!
I'd like to know what people working on machine intelligence plan to have their computers think about and if they're really trying to get them to think like humans.