| BitVyper |
03-25-2009 10:42 PM |
Quote:
it would essentially be no different from how we currently train animals for purposes we find them useful for.
|
Funny thing about that: You can train and train and train as much as you want, but an animal will always have a mind of its own. Oh, you can get them to perform their tasks flawlessly, but that doesn't mean they stop thinking about other things. It's just that it's really easy for us to keep control over their circumstances and provide them with the things they want and need. The world's working dogs aren't about to form a union. Depending on the complexity of the AI, that may not necessarily be the case. If you create something that can handle abstracts for some reason (maybe it needs to deal with people, or problems that require lateral thinking), then maybe it can start to desire and think about abstracts in ways you didn't predict. Of course, we're a loooooooooooooooo- 3 hours later -oooooooooooong way from having to be concerned about our AIs being too advanced. Hell, if we COULD produce animal-level intelligences in our computers, it would be absolutely gorram incredible. Like science fiction incredible.
|