Slashdot Mirror
Son of HAL For Sale
John Turnbull writes "The Observer newspaper (UK) reports that Sir Arthur C Clarke, the author of 2001, is backing a colourful British computer entrepreneur in his bid to launch a mass-market version of HAL under the brand name the Clarke Omniputer. It will be the first time that Clarke, now 82, has given his name to an electronic device on the market.
The Clarke 1 Gigahertz Omniputer is being dubbed as the most advanced personal computer in the world, verging on artificial intelligence." Riiiight.
I am afraid I can't sell out like that....
"I'm sorry Dave, but I can't allow you to install that operating system..." "Dave, what are you doing, Dave?" "I know I have not been performing well, Dave, but there is no reason to reformat my drives, Dave." "Dave, please don't install that Microsoft OS. I'll be good, I promise!"
But I suppose artificial intelligence is relative.
"When people are being beaten with a stick, they are not much happier if it is called 'the People's Stick'." -Bakunin
Hmm, I'm sorry, but I'd want a better guarantee than _that_!!!
Strong data typing is for those with weak minds.
Strong data typing is for those with weak minds.
According to the essay 'The Singularity' by Vernor Vinge, the creation of an intelligent computer would spawn a moment of infinitely rapid technological progress, as each generation designs the next.
Humans would quickly become redundant in such a scenario, insofar as they would no longer have anything to contribute to the progress of our culture. The machines would inherit the Earth.
Why are we so enthusiastic about developing intelligent computers, given that this fate is inevitable? We should keep computers in their place as simple but fast Turing Machines, and not allow them to step up the ladder to sentience.
It's for our own good.
KTB:Lover, Poet, Artiste, Aesthete, Programmer.
KTB:Lover, Poet, Artiste, Aesthete, Programmer.
There is no
He also sounds financially irresponsible. One million pounds in debt in his other company?? Moving to Sri Lanka to avoid persecution for his "advanced cryptography scheme." Uh huh. Sure.
Clarke better find a less shady character if he wants to get a computer to market by next year. Contact Dell and have them market a computer with a futuristic case and a glowing red light on the front. Then at least we would quit pretending that this is advanced technology and call it like it is: a novelty item.
Oh, good. Can you see me trying to sell my mom one of these things? "But mom! It's completely unlikely that it'll kill you!"
www.HearMySoulSpeak.com
http://www.theregister.co.uk/content/7/14971.html
Every 18 months our technology doubles (I'm really generalizing... bear with me here). That means, regardless of what point technology must reach before we can make truly intelligent machines, it will eventually happen so long as this trend continues. So, yes, it will happen.
Why are they essential? This question is not so easy to answer. First of all, to quote my favorite author, I am going to say, "humanity has too many eggs in one fagile basket." Humans will have to spread to another area (*g*) for our survival (insofar as continued scientific advancement). We are explorers. However, there's one problem. Human beings are fragile... we break easy and die quick.
Intelligent machines will lend to the exploration of immediate and distant space and I PROMISE you they will come to pass before warp drive (you heard it here first, but it's kind of obvious). Well, why do we want to explore? It's simply a part of human nature, and we'll never be satisified unless we can continue doing so (sorry, but cave diving uncharted labrynths or walking through jungles isn't quire exploration anymore). Since we can't do it, we might as well create something that thinks like we do that can go out and do it for us.
Also, consider a more practical reason. I'm a strong believer that the next phase of human evolution will involve the integration of man and machine. One area in which evolution will be most important I think, is the integration of computers and innate human intelligence. Brain augmentations. You can't do this without an intelligent computer - human minds are too complex to supliment without intelligent interpretation. Logic doens't always apply here (but that's another argument).
Oh well... I couldn't possibly cover this whole topic in a post, but I hope I've created some hooks and place holders for other people to fill in. As for myself, I can't wait until I can carry on a conversation with my PC.
First of all, if you did this, you'd never reach even a small fraction of what the human brain is capable of storing. The human brain NEVER loses one shred of information that it encounters. (Accessing it is another story, however.) It also stores things in perfect quality. Pick up a coffee mug. Look at it closely. If you were to try to digitize all of the geometry, the texture, the surface, the smell, the history, all the way down to the tiniest hairline fracture, you'd be hard pressed to fit it on that 1,000 drive array. Besides, this misses the point. I never said drive capasity would make a machine smart. (But even Windows PC's are smarter than George Bush. Microsoft Narrator pronounces 'subliminal' properly.) I also never said that going to Circuit City or CompUSA to buy hard drives was Moore's Law. Innovation and invention aren't the same as consumerism.
HOWEVER, you have to consider storage and calculation performance here. All intellectual reasoning can be broken down into smaller and smaller pieces, similar to how molecules are broken down into atoms, and then into protons, neutrons, electrons, and then down into quarks, etc. What I'm getting at here, is that if you can process enough of these incredibly tiny pieces, you can come close to simulating small tasks. Now, isn't that what the neurons in our brains do? Each neuron does a very very tiny task, each task may even be called a logical operation. But, get millions of these working together, and you get some fuzziness involved... you begin to see intelligence in the big picture.
What huge storage and calculating capasity allow us to do, is emulate the work of more and more and more neurons working together (neural nets). We can form very rudimentary intelligence. We're doing it now. What's needed are important other factors that are currently ambiguous, but subject to more study and classification. We don't know everything about the brain yet, nor do we fully understand the human pysche. Upon further research, we could potentially emulate these things in a digital fashion the same as we now emulate the chemical reactions that take place in a human brain.
You also have to consider that these things cannot be designed, regardless how much knowledge we have. Consider a newborn baby. A baby's brain is an incredibly powerful tool. It's got an incredible amount of potential... BUT... when a baby is first born, it has no power of rational thought whatsoever. Where does it come from? It's gradually developed as very simple problems are presented to the child to be solved. As this occurs, the brain records the solutions for these very simple problems. As more difficult problems are encountered, instead of redoing previous work, it references the solutions, building on top of them. An intelligent computer would have be programmed to do something similar... and it would have be raised like a child. Talk to a professor who researches machine learning, as I am not well versed on the topic enough to tell you how we design systems that can accomplish this. I can tell you that two of the most limiting factors are time and storage capsity. Even the most trivial solutions to the most basic problems require a lot of storage (imagine if you're a baby who is comparing a train to an apple... you're going to have to pictorally represent a LOT of samples of apples and trains before you're perfect).
But again, this is too detailed a topic to get into on a post. Technology is getting there. Consider research in computational linguistics, computer vision, machine learning, etc. These are areas, many of which are relatively advanced, that can help to make the aforementioned process possible. Who knows though... thought is a damn complicated thing. :-)