I finished reading Ray Kurzweil's 1999 book "The Age of Spiritual Machines -- When Computers Exceed Human Intelligence". It was an incredibly fascinating read. Way too much to try to explain everything here. I'll do some exploratory posts one small topic at a time.
The basic premise behind this futurists predictions involves the "Law of Accelerating Returns" -- which is basically that our technology is growing exponentially -- and that our next phase of evolution will be merging with our technology. We are already experiencing this in some small ways -- especially in the areas of genetics, artificial body parts and information/communication.
Ultimately, as we start using technology to enhance our physical and cognitive abilities, and as computers get smarter and start behaving more like humans, the line between human and machine will become increasingly fuzzy. By the end of the century, Kurzweil predicts, that merger will become complete, and those humans that have choosen to remain in their original state will not be able to participate in a meaningful way.
One question that Kurzweil couldn't answer involves the following:
Imagine that you have a computer powerful enough to fully simulate a human brain -- and a way of scanning someone's brain to get a complete, detailed, accurate model of all its intricate connections. (Say by the year 2040). If you downloaded your mind to a computer, it would think and behave exactly like you. It would have all of your memories and experiences. It would claim to be you -- and people might be inclined to believe it -- were it not for this sticky situation that there is also a person walking around (the original you) that claims to be you.
So -- what really makes you, "you"? Are the set of patterns now on the computer conscious, and are they the same consciousness as the original you? Does that consciousness live in your brain, or is there something beyond the physical where that consciousness truly resides -- ie. your soul, spirit, the "collective consciousness of the universe" or whatever?
I have some of my own thoughts, but I'd like to hear yours first!
jbravo
14 Comments
It's dopey compared to me.
Your 'puter sucks to the power of 600,00099
One difference is that the computer brain would be operating at much higher speeds. The perception of time to this computer entity would be far different. A millenium could pass to this virtual being in 10 of our minutes. Thus, unless it had a virtual world to exist and interact in, it would quickly become bored and then go insane. I think this is something we need to be very careful about when we consider the implications of doing something like this.
-JB
Post more!
Where did you get this book???
Is it not true that we have only begun to discover the capabilities of the human brain?
How could we, in the short time of fourty odd years, replicate something as complicated as it is.
Anyway, I have to get this book. Any links?
As an example in his 1999 book, he cites the Human Genome Project. I think he said that 5 years into the 10 year project, they were only 1% complete. People asked the same questions as you. Yet, because of rapidly evolving technology, it was completed right on schedule.
You can probably pick it up at your local library. If not, here's an Amazon Link:
http://www.amazon.com/Age...oks&qid=1184282602&sr=8-3
I'm also getting ready to start his 2005 book. It will be interesting to see what new ideas he's come up with in 6 years:
The Singularity Is Near: When Humans Transcend Biology
http://www.amazon.com/Sin...oks&qid=1184282602&sr=8-1
Also, take a look at his web site: http://www.kurzweilai.net/index.html?flash=1
penumbra2000 said (in Science Shows Global Warning is a Scam):
"Although, the requirement for a scientific theory to be falsifiable is a questionable one. It eliminates doing real fundamental theory work in a number of important areas, for example the behaviour of the mind: there's a form of Heisenberg-style uncertainty at play, where the presence of experimentation changes the results."
I responded with:
"Your comment is very similar to something I was planning on saying there [What makes you, "you"], but hadn't done yet. And that is -- I think there is probably critical information in the brain that is stored as quantum state. This would solve the problem of having 2 "you"'s around. The scan to get all the required data to replicate your brain on a computer would destroy the quantum state of your brain in the process -- which I think is ultimately what defines who you are. Thus, your consciousness would essential be transferred to the computer, rather than duplicated. This would also apply in cases of teleportation."
Other examples include things like the 'authority effect' (I'm not sure of its clinical name), in which whether or not the person running the study wears a white lab coat and speaks clearly affects the results of many psychological studies; and the problem inherent in double-blind subjects correctly (or incorrectly) guessing what the study is about and adapting their responses accordingly.
I'm pretty sure you're right about the teleportation issue, btw--to transmit an object via a digital medium, information about the relative positions and velocities of all the component particles which make up the object would have to be known; the only way we know of to obtain this information is to collide the particles with other particles of roughly the same size, which of course significantly impacts the relative positions and velocities of the particles. Hence, in order to scan the original object to the level of detail necessary to perfectly replicate it, you must destroy it.
Although I can think of a couple of theoretical ways around this. Heisenberg's principle only stands in the face of current imaging techniques, I think, which all basically involve playing pool with subatomic particles and watching how they hit the bumpers. If some aspects of our standing quantum theory aren't accurate (a distinct possibility, given that quantum theory introduces at least one rather sticky logical circularity) it may be possible to read the state of a particle through passive means, in which case Heisenberg's limitation dissolves.
Another, more mechanical, solution could be to coordinate the samples in a massively, ridiculously organised manner (which is not as crazy as it may seem, as this kind of absurd level of coordination seems to be a requirement for teleportation in the first place). You could concievably displace every particle in the object being scanned in exactly the same way simultaneously; the object being scanned would go through some near-light-speed displacement (without any macroscopic inertial side-effects, like lungs colliding with rib cages and the like), but otherwise remain intact. Although this might also result in a very large emission of heat, possibly cooking the person alive. It might also be theoretically impossible; it might require you to know the relative positions and velocities of the object's constituent particles ;P
http://www.woyano.com/view/5283/Virtual-Sex?24014
The last theory on how the brain stores information that I heard about was the holographic model, which I don't entirely get. The major defining feature about a hologram is that, if you slice it in half, you get two smaller, lower-resolution copies of the whole; the 'structure' is a sort of field-effect which is smoothly distributed throughout the entirety of the hologram. This is a little like the old kind of gifs, which would load a few gigantic pixels, then resolve the big pixels into smaller ones, then those into smaller ones, until the image was finished. If you dropped the latter half of the file, you still had the whole image, but bigger pixels (or, looked at another way, a smaller image with the same size pixels).
Another thing I've heard about brain physiology is that it is consistently defying mapping. Physical parts of the brain can be generally mapped to physical activities, like parts of the body, visual decoding, some aspects of emotion, and so on, but 'the mind', whatever it is, doesn't want to be located. Also, physical brain functions seem rather portable, with miscellaneous parts of the brain taking over for damaged ones.
Another aspect of this question that I'm interested in, a more philosophical one, regards the definition of intelligence. I've heard quite a large number of functional definitions, which identify specific kinds of intelligence, or allow us to decide whether or not to call a particular machine 'intelligent' (the Turing test, among others). But I haven't seen an adequate definition of intelligence that I can live with at all.
So my big question is, how do we know that *we* are intelligent? It's virtually always a basic assumption, and people cite ad-hoc evidence as proofs, such as that we can have this conversation, that we can override our instinctual responses, that we made fire, and so on. And given any of the aforementioned definitions of intelligence, we always qualify (at least, some of us do--depending on the outlined criteria). But when approaching the question of 'artificial' intelligence, we pretty much always fuzz away the question by labelling it as a special thing ('artificial' intelligence), and defining it in terms of how it emulates this or that characteristic of human intelligence (where such characteristics are also typically ill-defined).
My intuition on the subject is that 'intelligence' is a complex continuum, with 'inanimate' objects at one extreme, God only knows what at the other, and ourselves somewhere in the middle. There are many facets to intelligence, supplying different abilities and such, but that one of the foundational facets is symbology--in fact, I'm not entirely sure that there is any meaningful distinction between the concepts of language, communication and intelligence.
It seems to me that any group of beings displays a form of intelligence, just as so-called 'individuals' do (I say 'so-called' because the human animal is actually a complex collective of life forms, each of which also betrays some characteristics of intelligence). The human race as a whole is an intelligent life form which demonstrates the presence of a personality; you can see its actions as they progress over thousands of years, most easily by comparing historical maps and the relationships between human languages.
As for whether or not the actions of the human-species-animal, or the earth-ecosphere-animal, or whatever-animal, are intentional, I raise the same question about people. You're born embedded in a society, with some genetic predispositions and biological predispositions; you are raised in an environment which shapes you as you grow. The motivating forces behind a person's actions can be traced back to things that happened to them in the past, combined with their personal reaction to those things. Although we perceive all of this to be evidence of intention in our actions, it seems to me that that is precisely what it *would* look like, whether or not there is some notion of intention in our actions. (For that matter, isn't the notion of 'intention' defined relative to the human condition? Perhaps 'intention' isn't the right target...)
In any case, I'm not sure. I'd like to see what would happen if we did some serious study of the lifeforms of planet Earth as a collective intelligence. I'd even like to see a solid metaphysical framework in which we could do such a study be proposed; I think we could learn a lot about ourselves through such a framework.
Intelligence, and reality in general, is a strange thing. The more you dive in to try to understand the details, the more illusive it becomes. Understanding comes from context. The whole is greater than the sum of the parts. Trying to understand the nature of things through reductionism will only give us some of the answers.
I think you're right -- intelligence (and in my view, consciousness) comes in a whole continuum. That would seem to also be the case as we break down the processes in the human mind.
I'm not sure how we test for intelligence either. Right now, I think the best we can do is make relative comparisons. Is consciousness any easier to define?
Ever read "The Global Brain" by Peter Russell? It deals with the notion of collective (human) intelligence.
Imagine you had a brain injury and lost some of your ability to remember things -- but were otherwise unaffected. Would you consider yourself in the aftermath to be the same person, or now somebody different?
Now, imagine an electronic implant that fully restored your ability. You seem to be the same person you always were. Are you?
OK -- here's the slippery slope. Imagine that you replace every aspect of your brain and nervous system one piece at a time. Do it one neuron at a time if it helps. After each step, can you say you are still "you"?
You see where I'm going with this. After enough steps, your entire brain and nervous system has been completely replaced with electronics -- and yet, you appear to be the same person. So, is what made you "you" your brain cells and connections, or is it just the pattern of intelligence imprinted on them?
What's the difference, then, if you simply download your pattern onto a capable computer. Isn't that still you?
Freaky, hunh?
Hey you know AdGuy always gets the last word! ;)