Trust me, Will Robinson...Danger, Danger!
Posted on 06/12/2006 6:18:24 PM PDT by annie laurie
Ping
|
Tagline being fixed.
No problem - just yell "klaatu barada nikto!" and they'll heel.
And really hot infiltrators!
Trust me, Will Robinson...Danger, Danger!
"However, the idea that general-purpose robots, capable of learning, will become widespread is wrong"
I believe that sometime in the future artificial intelligence will increase exponetially to where a robot would learn very quickly, even human mannerisms. Also there is nothing stopping robot manufacturers of using wireless to allow them to tap into vast resources on how to do something they've never done before. In essence I'm saying that the robot in "I, Robot" is a very real possibility.
File that statement with this one...
"Everything that can be invented has been invented." --Charles H. Duell, Commissioner, U.S. Office of Patents, 1899.
[[I believe that sometime in the future artificial intelligence will increase exponetially to where a robot would learn very quickly, even human mannerisms. Also there is nothing stopping robot manufacturers of using wireless to allow them to tap into vast resources on how to do something they've never done before. In essence I'm saying that the robot in "I, Robot" is a very real possibility.]]
I agree! My argument and conjecture has always been, that once machines become self-aware (just before or just after) their learning capabilities will become incredibly exponential in mere seconds. Their cognitive systems won't be limited by time and distraction as humans' are. Of course, I have no idea what type of resources or "motivation" this supposed Robot will have - if any. My best guess is something "Borg"-like in needs. However, the Robot will not have millions of years of developed survival instincts. I think it'll be easy to destroy or "unplug", if necessary.... sort of like a smart version of a Berkley college professor (without tenure).
as my old machining instructor was fond of saying:
you ALWAYS turn the machine off.
you ALWAYS flip the lock-out cover plate into place.
you ALWAYS insert the red safety pin/flag.
you NEVER trust anyone, and you NEVER assume a dad-blamed thing.
"My argument and conjecture has always been, that once machines become self-aware (just before or just after) their learning capabilities will become incredibly exponential in mere seconds."
Can you elaborate on this for me? I've done a lot of thinking on vastly improved AI in robots and slowly I have built an imaginary robot in my mind into a very sophisticated "machine" that essentially is a great servant to all of mankind. You're talking about a great leap forward that I have not thought of. Once I understand your thoughts on this I think I can move forward faster.
Hi JwhDenver
I'm sure you know WAYYY more about it than I do. I'm using creative thinking and imagination based on what I THINK I know about AI, upcoming.
I imagine a combo electro-chemical processing with virtually instantanous speed. Now, such an explosion of self-awareness, followed by a RELATIVELY godlike knowledge - requires a lot of obvious prerequisites by the designers. Such an exponential processing event could move in millions of defined, finite, segments (one purposeful segment of knowledge, the "Joshua" theory) - or be dynamic and broad (the "Lawnmower Man" theory).... or, result in a locking loop that crashes the whole matrix (brain, or system) down.
So, take for example, the super chem system I describe above - and suppose a billion logic processes were activated (i.e., like scientists examining all those lines of DNA code by hand, taking years). The chembot extrapolates the billion, or million logic query processes and (by PURE luck), becomes self-aware (to a degree the techs standing nearby wouldn't even realize). The system will or won't do any number of things for itself. This, then goes to my "Borg" theory... my best guess, the chembot would likely acquire information to one driven end or path. I doubt that the techs witnessing all of this will ever have a chance to learn what that path is (or was).
Notice I use a lot of science FICTION in my examples above - movies? I'm a Star Trek hack who just thinks about stuff like this.
PS - I'm in Denver, too! Best, Bill
ping
"I'm sure you know WAYYY more about it than I do."
LOL! No way, my idea probably got hatched when I was sitting on the can! Haha. I doing what you're doing:
"I'm using creative thinking and imagination based on what I THINK I know about AI, upcoming."
I get your idea and I think you're right if suddenly an explosion went off into a robot's mind you'd have possibly a complete override of any programming in it. The new logic could/would render it stupid and disregard it. Like you said in seconds you'd have a totally different robot which would establish it's own direction whether good or bad. BTW, I just ordered "Lawnmower Man" from Netflix, looks like a "thinking" movie. Thanks for the tip. What would the robot be in 1 day? Total unknown robot?
My approach was much more linear. My robot at first was simply a get drinks etc, relatively stupid robot. But as the company who made the robot kept upgrading the AI, I began to teach it how to interact with people and do many other things. All the while it also served as a walking computer and was able to give probabilities which came from already established computer programs via wireless. Then a new upgrade was installed and I gave it the name "Charlie" because it no longer acted like a stupid machine. It became smart enough to help figure out problems and work with humans to gain insight into wisdom. It learned how to use the vast information available to it and come up with various ideas that it would present either to me or other humans. Charlie became aware at some point during this time. As Charlie learned things he would store them in a "public domain" so other robots could access the learning and be able to do more and more things.
At one point robots were needed to explore some caves and so experienced mountain climbers were called in to teach some robots how to do it. On their first trip into a cave when they came to a cliff or whatever obstacle the mountain climbers would give them advice on how to get passed the obstacle. All this learning now became accessible in dealing with the next obstacle. It didn't take long for them to become very proficient at their job.
As the various robots became more knowledgeable the learning was linear, then slowly went to geometric and then parabolic because all information was stored in this "public domain" storage unit. If one learned something it became available to all. Since I had and others had taught Charlie he was the most advanced and set down the programming essentially as thinking servants to humans.
Charlie then used massive computers to improve his AI and spent a great deal of time programming various sectors of his brain to greatly expand his thinking capabilties. Each program was massive and when he finally turned them all on he was instantly completely overwhelmed with information overload. Way too many options. So he cut down each program to about 2 or 3% to alleviate this problem. As time went on he established a system to block worthless ideas before they ever came to mind. As time went on he turned his AI programs up a bit at a time. I haven't got to the point where he has them full on.
Years ago when I had a Commodore sitting on my desk I had some friends over and everyone of them worked with computers. One was in IT, another was proficient programmer, another had used them extensively in his work, and a couple of others who used them everyday. I told them one day computers would not have hard drives, it would all be ram. We were friends so they were easy on me in telling me I was FOS. 3 days ago I saw a major manufacturer came out with a laptop based on flash memory, no hard drives.
Denver A? I now live in Parker.
This is a true gas, 2 people using imagination projecting into a might be future without hindering ourselves with what is now.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.
