Originally posted by Liberator
Well like, it would find faults in itself and just build on them, improve them beyond that of which the scientists thought capable.I don't know much about computers so I'm most likely wrong.
Actually that's the idea behind the "hard" singularity. You construct a computer that can build a computer smarter than itself and it does so. Than that computer builds a computer smarter than itself.
Repeat a few hundred times and you have a techno-god.
Originally posted by Colossus-Big CIs it possible for a computer to gain self awareness?
With any amount of technology, is it possible?
I would venture and say "no." Nothing physical is capable of consciousness. Yes, technology (code) exists to program machines and computers to act upon stimulus/algorithms, but they will never understand why. Machines and computers simply do the bidding (without question). Never will a machine or computer ask for a vacation, speak of personal rights (dignity) and/or feel grief over the death of a loved one, for example in the truest sense.
"Why should a bunch of atoms have thinking ability? Why should I, even as I write now, be able to reflect on what I am doing and why should you, even as you read now, be able to ponder my points, agreeing or disagreeing, with pleasure or pain, deciding to refute me or deciding that I am just not worth the effort? No one, certainly not the Darwinian as such, seems to have any answer to this.... The point is that there is no scientific answer." -Darwinist philosopher, Michael Ruse
Originally posted by ushomefree
Never will a machine or computer ask for a vacation, speak of personal rights (dignity)
Categorically untrue. I'm terrible with computers and even I can write a VB program that can ask for vacation time and demand that it has personal rights.
Originally posted by ushomefree
and/or feel grief over the death of a loved one, for example in the truest sense.
This is a better challenge but it raises the question of how you know that *people* feel grief or any other emotion. If the answer is that they cry or act sad or anything along those lines then a computer/robot can absolutely be built that would convince your it felt grief (but probably not in the next 50 years).
Symmetric Chaos-
Robots have already been developed in Japan (and I'm sure other parts of the world) that mimic human behavior, but these "human behaviors," are completely artificial. These robots, merely respond to stimulus dictated by the programs written for them - nothing conscience about that! Don't you understand, that machines, computers and robots only "behave" a certain way, due to programs/software written for them? That's not consciousness. How do you make the leap?
Originally posted by ushomefree
Symmetric Chaos-Robots have already been developed in Japan (and I'm sure other parts of the world) that mimic human behavior, but these "human behaviors," are completely artificial. These robots, merely respond to stimulus dictated by the programs written for them - nothing conscience about that! Don't you understand, that machines, computers and robots only "behave" a certain way, due to programs/software written for them? That's not consciousness. How do you make the leap?
I don't see how this is necessarily different than humans. All emotions are a response to stimulus, unless you tend to start laughing or crying for no reason.
And what is it that makes you think *people* really have feelings? How do you prove that another person is feeling sad if you throw out the appearance of sadness?
Yes, it is true; human beings do respond to stimulus, but we understand the stimulus (and/or make judgements about the stimulus affecting us on a MORAL BASIS)! In other words, human beings have "dignity," and therefore, for example, become angry if miss treated. We ask the question, "How dare you do that to me?" That's consciousness!! I mean, people commit suicide! Shouldn't that be enough to convince you that human beings truly hurt at their core - enough to actually kill themselves! That is true pain. A computer will never self destruct, ha ha! It's not aware in and of itself. Computers lack the ability to makes judgements upon themselves even; human beings can (and do). Hmm... I hope that helped convey my point more specifically.
Originally posted by ushomefree
Yes, it is true; human beings do respond to stimulus, but we understand the stimulus (and/or make judgements about the stimulus affecting us on a MORAL BASIS)! In other words, human beings have "dignity," and therefore, for example, become angry if miss treated. We ask the question, "How dare you do that to me?" That's consciousness!! I mean, people commit suicide! Shouldn't that be enough to convince you that human beings truly hurt at their core - enough to actually kill themselves! That is true pain. A computer will never self destruct, ha ha! It's not aware in and of itself. Computers lack the ability to makes judgements upon themselves even; human beings can (and do). Hmm... I hope that helped convey my point more specifically.
A computer can be programmed to make certain judgments about the information it takes in before acting. A simple moral code would be easy enough to write, just not very useful, strict utilitarianism in particular lends itself to being turned into an algorithm.
A computer can be programmed to kill itself after a certain stimulus, people just don't program that way because it's a waste of resources.
Most of the posters here don't even know what the word "quantum" means in the phrase "quantum computer". A soul doesn't define self-realization. Many creatures are self-aware but do not posses souls. There is no "proof" and no "light show" for those who try to pseudo-intellectualize the concept. It is that way with purpose, so that there will never be "Valerie 23s" in our (human) existence. To dream is wonderful ... to visualize the trouble that comes with the dream is even more profound. But it's only profound because it is so far out of our reach.
Originally posted by Symmetric Chaos
This is a better challenge but it raises the question of how you know that *people* feel grief or any other emotion. If the answer is that they cry or act sad or anything along those lines then a computer/robot can absolutely be built that would convince your it felt grief (but probably not in the next 50 years).
yes, but with neuroimaging we can see emotions that people aren't outwardly expressing
whether a robot can be made to simulate human emotion is a radically different question from whether they can be made to experience human (or something similar) emotion, no?
Originally posted by ushomefree
"Why should a bunch of atoms have thinking ability? Why should I, even as I write now, be able to reflect on what I am doing and why should you, even as you read now, be able to ponder my points, agreeing or disagreeing, with pleasure or pain, deciding to refute me or deciding that I am just not worth the effort? No one, certainly not the Darwinian as such, seems to have any answer to this.... The point is that there is no scientific answer." -Darwinist philosopher, Michael Ruse
someone hasn't kept up with their cognitive psych/neuro litterature, Mr. Ruse
Originally posted by ushomefreeIt's like seeing someone argue that computers work according to behaviorist psychology while humans work according to cognitive psych, with a deeper inner processing that allows humans to achieve dignity and 'true' emotion.
Yes, it is true; human beings do respond to stimulus, but we understand the stimulus (and/or make judgements about the stimulus affecting us on a MORAL BASIS)! In other words, human beings have "dignity," and therefore, for example, become angry if miss treated. We ask the question, "How dare you do that to me?" That's consciousness!! I mean, people commit suicide! Shouldn't that be enough to convince you that human beings truly hurt at their core - enough to actually kill themselves! That is true pain. A computer will never self destruct, ha ha! It's not aware in and of itself. Computers lack the ability to makes judgements upon themselves even; human beings can (and do). Hmm... I hope that helped convey my point more specifically.
Re: Re: Re: is it possible for a computer to gain self awareness
Originally posted by inimalist
yeseven if that is true, every aspect of what is called "self-awareness" has some correlate to neurological activity. If there is a soul, it has little to do with awareness, and would not prevent computers from being self aware
this isn't true (except for us not having the know how to make an aware robot now) for potentially 2 reasons
the first is that there are no quantum interactions at the neurological level that have a serious impact on awareness. Penrose's ideas are pretty much mythology in both the physics and neuroscience community, and while some quantum effects might mediate ion channels in the neurons themselves, there is absolutly no reason that a quantum computer would be necessary, because there are no quantum phenomena that need to be accounted for
the second is that understanding awareness is not simply a problem of not having enough power. Human consciousness is based upon the interconnected nature of our neurology, and its constantly changing interconnectivity. At this point, it might be more accurate to describe the problem as being one of cracking the neuro-code, or how patterns of activation represent coherent states of awareness, rather than just needing something with 10000x the power.
We would need a significantly more powerful computer if we ever wanted to simulate human awareness through an artifical brain (with billions of artificial neurons), and that specifically might be made easier with a quantum computer, but with the ever constant advances in micro-processing, a super-computer of some kind may also prove just as useful. Even then, we need more understanding of the higher areas (Frontal and pre-frontal regions, parietal lobe, some of the temporal-cortical pathways) before simulating them is going to provide much more data than fMRI anyways.
This also assumes that we are building awareness in robots in the very same way it occurs in humans. Awareness, as humans describe it, is most likely an epiphenomenon of our linguistic ability and intense social interaction. For this reason, it is possible to say that our brain is not built specifically for "awareness", but is aware because of some of the things it was built for. Because of this, there are some weird experiments (optical illusions being the simplest example) where our awareness is terribly ineffecient. It might even be the case that, when building aware robots, we need to avoid the human model, insofar as that is possible, to the end that our robots have an elevated sense of awareness than we do.
However, since our concept of awareness is inherently anthrocentric, it is debateable as to whether that would qualify as awareness, or just something new entirely
Dude, I totally thought up an outline for a P-Zombie that's practically real AI. I'm not even kidding. I briefly described it to Bardock. There's a very real reason why something like my outline will not be programmed: it would take an incredible amount of people and time to do. I estimated over 100 billion lines of code (almost an arbitrary number but I used an existing "AI" program as a jumping board to estimate my full program size) required to write out the entirely of the program. The program would still have to be "cleaned" and "tested" to such an extant that it's pretty much impossible to complete without, literally, hundreds of thousands of programmers/debuggers.
Basically, it covers the very same items you mentioned and many others.
We have "technology" now to write and execute a program like that, with no problem. The problem is amassing the data for all of the objects (I created 5 tiers of objects: Hyper objects, Super Objects, Macro Objects, Sub-Objects, and Micro Objects (these are words I made up but it's easy to see it's a pyramid)) and creating those objects to work together, correctly. The method by which they interact would be very simple but the objects themselves are what have to be "figured out" first.
I COULD write it all out, one day. It's much too complicated and lengthy to type out so I prefer to speak about in detail instead of type it out.