Moderators: carlson1, Charles L. Cotton
The stage has been set for competitors to vie for a $2 million prize from the Department of Defense to develop a robot that could perform a number of physical tasks that might be required to respond to a disaster or an emergency as part of the Defense Advanced Research Projects Agency's Robotics Challenge, which DARPA unveiled last October.
Never (maybe not in the next 200 yrs), can be simulated but it is still a dump machine and as good as its programmer. As Specialist and expert in the field myself, I do not discount the "close but no cigar" for those android. Biggest problem is that programming AI to do self-learning is really very tough.anygunanywhere wrote:When is self realization going to occur?
Anygunanywhere
BigGuy wrote:I remember back in the 60's when I first heard about digital computers. Futurists predicted that within 20 years, machine intelligence would match human intelligence.
Then in the 70's there were fantastic breakthroughs in computer science in the form of circuit integration and the advent of "computers on a chip." Futurists predicted that these new developments would lead to human level intelligence within 20 years.
Then in the 80's Scientists began studying the brain and these new insights into how the human brain actually works lead to predictions that we could match human intelligence with computers within 20 years.
In the 90's the concept of neural networking was discovered. And this process is so startlingly similar to the way the human brain works that computer scientists predicted being able to mimic the human brain within 20 years.
By the year 2000, demonstrator circuits using light and quantum mechanics were being built in the lab. Most researchers believed that this put us within 20 years of human level intelligence.
Then came Watson.
The Science behind Watson
[youtube]http://youtube.com/watch?v=DywO4zksfXw[/youtube]
What the video doesn't mention is that the algorithms that Watson uses were not written by programers. It developed it's own algorithms by analyzing previous correct answers. Some of Watson's developers think they will be able to match the human brain within 10 years
While I don't think it will be 10 years, I don't really think it will be 200 years either. We're in the early stages of machines that design, build and program themselves. The geek in me can't help but be excited by the idea. But somewhere in the back of my head is a nagging little voice asking it this is really such a good idea.
I don't think that these "Artificial Intelligence" entities will "match" the human brain. They will probably operate differently. But they may well be as, or even more, powerful than the best human minds. If we keep going along this path, Mike will wake up. No telling whether it will help the Loonies, the Warden, or just declare itself a free agent.
