Quote:
Originally Posted by LoneAspen
Having been in the IT world for about 25 years, much of writing code, my own personal feeling is that the truly intelligent, self-aware, sentient machine will never happen.
In fact, I think you could give the smartest people on the planet regarding AI (artificial intelligence) access to an unlimited number of infinitely fast processors, and an unlimited amount of infinitely fast memory and storage, and I still don't see it happening.
A sentient machine would have to be "bootstrapped" by something, and I think that something would have to be our own knowledge of how we as humans think and are sentient. But we don't know enough of all that to reproduce it (short of biological reproduction, ie, having a baby).
I would be willing to bet even another 2000 years from now, assuming technology continues to advance, we still wouldn't see a truly sentient machine. I just don't see it happening. Ever.
|
I'm not so sure about that. I'm a programmer/engineer - and know that the stuff I'm personally working on doesn't fit the bill in any way shape or form. Nor should it.
But I chaperoned my FLL robotics team to a field trip the the Neurosciences Institute in La Jolla. Those folks are approaching the programming in a completely different way. They're trying to learn how the brain works and are simulating it with a completely different style of programming. I didn't understand all of it - but it was pretty mind blowing. (We spent more time talking to the guys who built their crazy cool robots that play soccer - on segway platforms. So I didn't get a chance to dig deeper into it.)
There are people working on a whole new paradigm - much closer to the way the brain operates... as a way of confirming how the brain operates. Not just decision trees - but preferences, biases, some learned preferences, but others random and innate, etc. (Why do some people like chocolate, and others prefer candy corn?)
Synthetic Neural Modeling - The Neurosciences Institute