Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
I am not an AI specialist but I am a software developer with an interest in AI and I would say you've hit the nail pretty squarely on the head.
There have been very interesting strides in understanding how the brain works in the past few years. There is an open source library written in the Python programming language that does a respectable job of simulating the human prefrontal cortex. We know what each cellular layer of that part of the brain is responsible for, how the basic feedback mechanisms work, and the interesting thing is that the software model has actually been useful in figuring out how the brain works because when it produces results at variance with an actual brain, it can be fiddled with until the outputs / responses match what is known from studying the actual brain.
One thing that has hampered progress in AI until now is that we have made the fundamental mistake of assuming that the human brain is a glorified digital (or perhaps analog) computer. It is not organized along those lines at all. And what should have tipped us off to that is that although the brain's "clock speed" is a tiny fraction of that of our best computers, our best computers have utterly failed to mimic its most rudimentary capabilities such as determining on the fly what physical movements must obtain in order to catch a ball being thrown at it. Somehow the brain does that perfectly with a bit of practice, whereas even clumsily approximating that feat would require the combined resources of our best supercomputer.
The secret appears to be that the brain is mostly a pattern-matching engine. The best AI work I have seen of late approaches it from this angle, with much better success.
All that said, we have a LONG way to go and we still don't know how learned patterns are stored, retrieved, and made integrated use of in the brain. Likely any computer successfully dedicated to either simulating or reproducing consciousness would be built very differently from current machines, even if we understood how to replicate or transfer the information out of a living brain. And it begs the question, if we could transfer it from a living brain to a computer, why not from one living brain to another? That form of biological immortality would probably involve growing fresh "blanks" to move into, they would likely be far better than non-biological media.
Here's the simple way to look at it: our best computers still experience the Blue Screen of Death, kernel panics, and various mystery lockups, and that's just running conventional software. I don't see such an environment as a refuge from mortality anytime soon. Heck, I can't even get my DVR to function correctly with a wired Internet connection!
There are things on the horizon like biological computing and quantum computing that could move us forward rapidly, but I can't see us being within less than a hundred years of Kurzweil's vision. Kurzweil's fundamental error is equating the availability of a given quantity of computing POWER with adequate knowledge of how to HARNESS and DIRECT that power, with sufficient reliability, and with an adequate understanding of the problem domain itself. The truth is that the computing power we have is, metaphorically speaking, all dressed up with nowhere to go when it comes to hosting human consciousness.
Given that consciousness / self awareness is probably an emergent property of a particular psycho-biological system, my guess is also that it will be difficult if not inherently impossible to corral it and send it down a wire to be reassembled elsewhere. It would be like scanning the sum total of human knowledge and transferring its morality someplace else. Morality has no meaning apart from society as currently expressed, and self awareness has no meaning apart from one's being as currently expressed. It is not a "thing" or particular pattern of inconvenienced electrons, like an email, that can be usefully copied.
Please never stop posting...and I mean never in the Singularity-enabled sense.
You're right. He just violated his own tautology. But given your knowledge of reality, you're wrong. You're right, therefore, you're wrong. I play by Mystic PhD's rules, FYI.
You clearly haven't a clue about MysticPhD's views and my rules are the rules of science.
The fact that I reference the masses is the part you either missed or refuse to understand. Any thought that the masses have an accurate knowledge of our reality is nonsensical. Apparently for you in Rio Linda I needed to phrase it differently. "The problem is the masses' knowledge of what reality actually is usually does not track with what we DO know about it. It is rather based on some common sense model based on our sensory experiences of it." I will try to remember to simplify my posts further.
I agree with you that 'the masses' do not understand reality. Perhaps nobody really understands reality - not even science, but they do understand it better than the masses and avoid the 'common -sense' model.
Thus, when we talk of the God -question, the masses' commonsense reality model is pretty irrelevant and the better understood reality model of science is the relevant one.
However, Mystic, your model of reality includes a large amount of speculation and Faith: specifically the view that Revelation can give you reliable insights into reality that mere science cannot reach. This is a common claim of religion and it has no sound scientific or logical basis. I know this, as it has become evident over month or years of discussion with you.
Your theory is, I have to opine, not science, but faith -based speculation and I have to say that the general inability of those who look at your theory to find it convincing is scientifically and logically sound, even though you do commonly suggest that they do not understand 'Reality' and are using a 'commonsense' misunderstanding of it.
To put the matter straight, they have it right and you old chum, are being unfair (to pit it mildly) to them, apparently in hopes to discredit their arguments and to keep your own view of reality looking credible.
Now we have clarified that matter a bit, Mystic v. Marcinkiewitz, round two!
The fact that I reference the masses is the part you either missed or refuse to understand. Any thought that the masses have an accurate knowledge of our reality is nonsensical. Apparently for you in Rio Linda I needed to phrase it differently. "The problem is the masses' knowledge of what reality actually is usually does not track with what we DO know about it. It is rather based on some common sense model based on our sensory experiences of it." I will try to remember to simplify my posts further.
You are right, the "masses" generally postulate some unknown and empirically unknowable entity to explain things they don't understand and to provide a backstory, a deeper meaning for their internal emotional experiences.
You clearly haven't a clue about MysticPhD's views and my rules are the rules of science.
The rules of science declare that you need to evidence your claims - that they be testable - that they make predictions that you can test and verify - and that your claims be falsifiable.
Nothing you have posted fits this construct. There is no evidence coming from you. No predictions let alone verified ones. No tests that have been verified. And since you hide all your claims behind a cloud of "dark matter" and other buzz words - nothing is falsifiable either - which is how you like it because then you get to claim no one is rebutting your views - because you have given nothing rebuttable.
So exactly what rule of science do you follow? Except for using Sciencey language on occasion?
Please never stop posting...and I mean never in the Singularity-enabled sense.
Quote:
Originally Posted by mordant
I am not an AI specialist but I am a software developer with an interest in AI and I would say you've hit the nail pretty squarely on the head.
There have been very interesting strides in understanding how the brain works in the past few years. There is an open source library written in the Python programming language that does a respectable job of simulating the human prefrontal cortex. We know what each cellular layer of that part of the brain is responsible for, how the basic feedback mechanisms work, and the interesting thing is that the software model has actually been useful in figuring out how the brain works because when it produces results at variance with an actual brain, it can be fiddled with until the outputs / responses match what is known from studying the actual brain.
One thing that has hampered progress in AI until now is that we have made the fundamental mistake of assuming that the human brain is a glorified digital (or perhaps analog) computer. It is not organized along those lines at all. And what should have tipped us off to that is that although the brain's "clock speed" is a tiny fraction of that of our best computers, our best computers have utterly failed to mimic its most rudimentary capabilities such as determining on the fly what physical movements must obtain in order to catch a ball being thrown at it. Somehow the brain does that perfectly with a bit of practice, whereas even clumsily approximating that feat would require the combined resources of our best supercomputer.
The secret appears to be that the brain is mostly a pattern-matching engine. The best AI work I have seen of late approaches it from this angle, with much better success.
All that said, we have a LONG way to go and we still don't know how learned patterns are stored, retrieved, and made integrated use of in the brain. Likely any computer successfully dedicated to either simulating or reproducing consciousness would be built very differently from current machines, even if we understood how to replicate or transfer the information out of a living brain. And it begs the question, if we could transfer it from a living brain to a computer, why not from one living brain to another? That form of biological immortality would probably involve growing fresh "blanks" to move into, they would likely be far better than non-biological media.
Here's the simple way to look at it: our best computers still experience the Blue Screen of Death, kernel panics, and various mystery lockups, and that's just running conventional software. I don't see such an environment as a refuge from mortality anytime soon. Heck, I can't even get my DVR to function correctly with a wired Internet connection!
There are things on the horizon like biological computing and quantum computing that could move us forward rapidly, but I can't see us being within less than a hundred years of Kurzweil's vision. Kurzweil's fundamental error is equating the availability of a given quantity of computing POWER with adequate knowledge of how to HARNESS and DIRECT that power, with sufficient reliability, and with an adequate understanding of the problem domain itself. The truth is that the computing power we have is, metaphorically speaking, all dressed up with nowhere to go when it comes to hosting human consciousness.
Given that consciousness / self awareness is probably an emergent property of a particular psycho-biological system, my guess is also that it will be difficult if not inherently impossible to corral it and send it down a wire to be reassembled elsewhere. It would be like scanning the sum total of human knowledge and transferring its morality someplace else. Morality has no meaning apart from society as currently expressed, and self awareness has no meaning apart from one's being as currently expressed. It is not a "thing" or particular pattern of inconvenienced electrons, like an email, that can be usefully copied.
You failed to mention "error" and "volume". our brain handles error far better. It is slower than a computer per electron route that is. But it gains speed through sheer volume of thought. I tend to agree with you in data transfer notion. It's like talking about resolution of the eye in mega pixels. It has no real direct correlation.
"wanting something more" is independent of there being something more. I am sure you know that the universe may be a quantum computer. Intellectually speaking, I can't say there is nothing more. In fact, the probability is that there must be something more.
That in no way suggest "bearded guy in the sky" pointing his figure at us. I have no idea what it would be.
I am not an AI specialist but I am a software developer with an interest in AI and I would say you've hit the nail pretty squarely on the head.
There have been very interesting strides in understanding how the brain works in the past few years. There is an open source library written in the Python programming language that does a respectable job of simulating the human prefrontal cortex. We know what each cellular layer of that part of the brain is responsible for, how the basic feedback mechanisms work, and the interesting thing is that the software model has actually been useful in figuring out how the brain works because when it produces results at variance with an actual brain, it can be fiddled with until the outputs / responses match what is known from studying the actual brain.
One thing that has hampered progress in AI until now is that we have made the fundamental mistake of assuming that the human brain is a glorified digital (or perhaps analog) computer. It is not organized along those lines at all. And what should have tipped us off to that is that although the brain's "clock speed" is a tiny fraction of that of our best computers, our best computers have utterly failed to mimic its most rudimentary capabilities such as determining on the fly what physical movements must obtain in order to catch a ball being thrown at it. Somehow the brain does that perfectly with a bit of practice, whereas even clumsily approximating that feat would require the combined resources of our best supercomputer.
The secret appears to be that the brain is mostly a pattern-matching engine. The best AI work I have seen of late approaches it from this angle, with much better success.
All that said, we have a LONG way to go and we still don't know how learned patterns are stored, retrieved, and made integrated use of in the brain. Likely any computer successfully dedicated to either simulating or reproducing consciousness would be built very differently from current machines, even if we understood how to replicate or transfer the information out of a living brain. And it begs the question, if we could transfer it from a living brain to a computer, why not from one living brain to another? That form of biological immortality would probably involve growing fresh "blanks" to move into, they would likely be far better than non-biological media.
Here's the simple way to look at it: our best computers still experience the Blue Screen of Death, kernel panics, and various mystery lockups, and that's just running conventional software. I don't see such an environment as a refuge from mortality anytime soon. Heck, I can't even get my DVR to function correctly with a wired Internet connection!
There are things on the horizon like biological computing and quantum computing that could move us forward rapidly, but I can't see us being within less than a hundred years of Kurzweil's vision. Kurzweil's fundamental error is equating the availability of a given quantity of computing POWER with adequate knowledge of how to HARNESS and DIRECT that power, with sufficient reliability, and with an adequate understanding of the problem domain itself. The truth is that the computing power we have is, metaphorically speaking, all dressed up with nowhere to go when it comes to hosting human consciousness.
Given that consciousness / self awareness is probably an emergent property of a particular psycho-biological system, my guess is also that it will be difficult if not inherently impossible to corral it and send it down a wire to be reassembled elsewhere. It would be like scanning the sum total of human knowledge and transferring its morality someplace else. Morality has no meaning apart from society as currently expressed, and self awareness has no meaning apart from one's being as currently expressed. It is not a "thing" or particular pattern of inconvenienced electrons, like an email, that can be usefully copied.
Rather than some euphemistic nonsense like "emergent property" . . . what consciousness has that a machine will never have is LIFE, mordant. That is why it is conscious. Only life can be conscious. Machines and circuits (no matter how elaborate and advanced) will never be alive. Therefor they will never possess consciousness.
Rather than some euphemistic nonsense like "emergent property" . . . what consciousness has that a machine will never have is LIFE, mordant. That is why it is conscious. Only life can be conscious. Machines and circuits (no matter how elaborate and advanced) will never be alive. Therefor they will never possess consciousness.
Now, that has got to be false reasoning. One may as well say that cloned sheep are impossible because sheep can only be made the usual way. While consciousness evolved along with life and indeed part of it, that does not mean to say that it cannot be duplicated artificially. To define it as something that isn't consciousness even if it is exactly like it, if it has not developed in the usual way is simply a cheat.
I'm surprised to see you do it.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
