re: #105 Obdicut

Um, we are not even close to proving that we could create ‘true ai’, unless you mean a biologically based one.

Where do you get the idea that we’re close to true AI?

We are very, very close. When I get home, i’ll post links to some of the work being done which I have saved there. There remain two essential problems left to solve: the availability of sufficient computing power, and the development of a functional cortical model. The first will happen within 20-30 years, the second will take a bit longer. There is nothing magical about intelligence or sentience, it’s just math. We know meat (brains) can do the math, and so too in time will computers.

re: #109 ArchangelMichael

“50 years away” is Scientist and Engineer-talk for vaporware that they have no idea if they will ever be able to figure out. Nuclear fusion power has been “50 years away” since the 1950s. It’s like asking a contractor when your room addition is going to be done and he says “2 weeks”.

Nuclear fusion hasn’t been reliably producing advances at an ever increasing rate. It isn’t vaporware, it’s inevitable, provided we don’t destroy ourselves first. The mathematical models already exist in pieces, it’s just that sufficient computing power to adequately test them doesn’t. That will change, and when we get closer to around 10^16 effective units of computation per second (the ballpark at which our brains operate) we will be able to test them more thoroughly.

Nuclear fusion is problematic because we don’t know how to do it reliably and controllably (and with a net output) outside the sun. Every vertebrate on this planet is carrying within it a working model of some form of intelligence. We will replicate it in time, I think, we just need time to do so.