The Singularity

[DavidV]

If we matched our computer power with the brain’s, how would AI suddenly have the proper design to allow it to surpass the brain?

It probably wouldn’t. Vinge & Co argue that computing power will have to significantly exceed the human mind first just for this reason. The necessary software will initially be far less efficient and capable than a human mind. Whether such software can be created is certainly a question, but we already have low-level use of AI to indicate that it is possible. The IP packets that this post was transmitted in were intelligently routed to your computer through one such application. Voice recognition, financial analysis, medical diagnosis, and the Mars Rovers also use AI. Even if a pure AI is not possible, integrating such systems with human intelligences can create a super-human intelligence.

If they are still human, they cannot be considered “super_human." If they are not human, they cannot be considered "super_human.”

The point is that accelerating technology will allow us to improve the design of our minds, and apply the law of accelerating returns to consciousness itself. This removes a fundamental limitation of evolution-bound consciousness, and as such is comparable to the initial change from non-life to life.

What such beings call themselves is besides the point.

4. Neuroscientists are studying things such as the “collective mind” or the “social mind” or the “altruistic mind.”

How is it that they will advance the understanding of the mind when they don’t understand it themselves (beyond a mass amount of particulars they have compiled)?

Such a generalization is decades behind the current state of neuroscience.

Drew1776 said that he has seen no evidence to suggest computers becoming as strong as the brain. Your response was to call him ignorant, and then support your claim about AI with the idea that a desktop computer can hold more binary data than a brain. Then you leap to saying computers will out-perform the human mind. Don’t you think you need a little more information to justify your claims?

My point was that Drew’s ignorance of the evidence is not evidence that something is impossible. It just means that he’s not qualified to conclude either way.

I can’t prove that AI is possible (no one can till it happens), but I can offer some reasons why it might be - the main one being that WE are rational machines.

As far as the Vinge essay showing that it’s all “SURE”, I don’t see how you think Vinge demonstrates anything with certainty when he doesn’t think so himself. Every step of the way he says “if” and “may” and "I think

What he said was that “If super-human intelligence is possible, then it will happen.” Then he some reasons why its possible. It’s not intended to be a bullet-proof argument - but the other sources I linked provide much more evidence.

God indeed.

You missed the point. The quote is not meant to be taken literally.

[Cogito]

What does this mean? My pocket calculator has more “power, speed, and utility” than me when it comes to arithmetic. Are you claiming that there is a special realm of computing restricted to biology and/or the human brain? What is the evidence for this?

There is a vital difference between computing power and consciousness that I think you’re missing… I don’t really want to go into the whole issue, but if you’re willing to read a book that has its faults but explains the issue well I reccomend The Emperor’s New Mind by Roger Penrose.

[DavidV]

There is a vital difference between computing power and consciousness that I think you’re missing… I don’t really want to go into the whole issue, but if you’re willing to read a book that has its faults but explains the issue well I recommend The Emperor’s New Mind by Roger Penrose.

I’ve heard of his theory of consciousness. It basically goes “well, quantum physics is a mystery, and consciousness is a mystery. They must be related! Here’s a theory I just dreamed up to prove it!” It’s just another attempt to defend volition by finding a “loophole” around causality.

[Cogito]

I’ve heard of his theory of consciousness. It basically goes “well, quantum physics is a mystery, and consciousness is a mystery. They must be related! Here’s a theory I just dreamed up to prove it!” It’s just another attempt to defend volition by finding a “loophole” around causality.

Ok, fine… Can you explain the connection between being able to compute vast sums using completely deterministic algorithms is related to volitional consciousness?

[mweiss]

I’m watching a prerecorded interview of Ray Kurzweil, broadcast on C-SPAN2 from November of this year. So far, in this 2+ hour interview, he’s made some interesting assertions. One of these is that quantum mechanics involves consciousness. He essentially expressed the primacy of consciousness as a fundamental element of quantum mechanics (ie., the observer affects the behavior of particles.) Years ago, I had heard of the experiments with Schroedinger’s Cat, and had a hard time with the concept of the cat influencing external devices without physical contact.

Ray Kurzweil is an extraordinarily fascinating individual. A look at his web site, http://www.kurzweilai.net, gives a small glimpse of the broad range of future technology that he is involved with. http://www.kurzweiltech.com actually encompasses all of the Kurzweil companies. It’s a vast world.

With regard to the exponential progression of computer power and passing the Turing Test (the ability of a computer “being” to interact with humans at a level where humans can’t tell that it’s a computer that they’re interacting with), Kurzweil asserts that technology is already being used to perpetuate more advanced technology. By amplifying the power of the human brain, technology can make possible the development of even greater technology. Indeed, many of the large scale integrated circuits in use today were produced with enabling technology that removed some of the limitations of human engineering.

Ray Kurzweil is working on life extention technology now. Apparently he considers death to be a tragedy in that the knowledge of a person—their experience over a lifetime—is lost when the person dies. He believes that the technology will extend life incrementally, where one particular problem will be solved, and it may gain us 5 years of life, and during those 5 years, more advances that will gain us 10 or 15 years of additional life will be developed. He predicts that in 20 years, we’ll have technology that will gain us one year of life per year of life lived.

With regard to computers attaining self-awareness, I don’t know how this will happen. What we tend to ask is that question of existence that man has pondered since the dawn of the intellectual era, is what makes man “spiritual”—what gives man a “soul” a sense of self-awareness, values, desires, etc. Whether a computer of the future can attain some level of “consciousness” remains to be seen.

Kurzweil’s predictions about future technology have been highly accurate so far. As such, I find his predictions fascinating.

[ExtremelyAmerican]

What does this mean? My pocket calculator has more “power, speed, and utility” than me when it comes to arithmetic. Are you claiming that there is a special realm of computing restricted to biology and/or the human brain? What is the evidence for this?

Yes. It’s called conceptualization. Read “Introduction to Objectivist Epistemology.”

In regards to what you view as an anachronistic generalization about neuroscience: I’d like to know what you think is a more accurate description of the modern field. I’d also like you to explain how these findings correlate with anything being done in AI or computer hardware. For instance, how do IP packets being routed to my network have anything to do with the human brain? More importantly, what do any of the latest advances in AI (usually intricate systems of identifying perceptual data within given parameters) have to do with the focus of modern neuroscience on curing and healing brains?

To my knowledge, AI scientists and neuroscientists don’t work together; why would they need to? Computers are designed to supplement our ability to hold perceptual data; check out the idea of “Crow Epistemology.” What besides this have you demonstrated is possible? What exists besides memory databanks with simple (or intricate) filing systems? Besides all that, what sign do you have that people in the field of AI (or neuroscience) have any idea of what the human mind can do and how it can do it?

Basically all I have seen from this conversation and the Vinge essay is extrapolation from what already exists into an imaginary, science-fiction-like world.

[edited typo]

[DavidV]

Conceptualizing is certainly something that no current software program can do. But you said that “human mind is beyond the power, speed, and utility of any computer.” How do you know that? It’s conceivable that computers in the future could conceptualize. They might be optical, quantum or biological - we don’t have any evidence to suggest that one platform is superior to any other.

If you accept causality and classical physics, then a computer should be able to simulate a brain in its entirety, so any modern computer should be able to run a simulation of the brain. (Albeit much slower than the real thing.)

There are already many supercomputers than exceed the brain in terms of raw processing capacity. The only question is whether we can write software that can conceptualize. This is certainly not a sure thing, but there is no reason to assume that it is impossible.

(Even if it is impossible, it has little bearing on the singularity idea. Human consciousness can be enhanced by biological means even if digital ones are not possible.)

Also, terms like “power, speed, and utility” are meaningless without a context. As I explained, very simple devices easily beat human consciousness in terms of “power, speed, and utility” when it comes to certain functions. The functions that computers are not capable of are not “inferior” to human ones - the comparison is simply not valid. It’s like saying that my vacuum is better in “power, speed, and utility” at cleaning carpets than my toaster.

To my knowledge, AI scientists and neuroscientists don’t work together; why would they need to?

These fields are so closely tied today, that scientists are practically required to be experts in both fields.

For examples of applications, see

IBM is creating a supercomputer simulation of the brain

http://en.wikipedia.org/wiki/Computational_Neuroscience

http://en.wikipedia.org/wiki/Brain-computer_interface

http://en.wikipedia.org/wiki/Neural_engineering

What besides this have you demonstrated is possible?

Actually you claimed that AI is impossible, so the burden of proof is on you. I already pointed out some reasons why it might be possible.

What exists besides memory data-banks with simple (or intricate) filing systems?

Well, for example, the vehicles that won the DARPA challenge last year have AI software that learns from its mistakes to learn how to “see” obstacles and drive without human guidance. Evolutionary algorithms use digital DNA to evolve their designs to better meet business requirements - such as picking the best stocks. Some viruses use polymorphic code to evade scanners, which use heuristics to detect new viruses not in their definition files. Spammers scramble their ads to evade anti-spam software which tries to decipher jumbled images to detect spam words dynamically created based on Bayesian rules from user feedback. This post was spell-checked using suggestions from a vocabulary dynamically created from users misspellings on Google. That’s a little more advanced than a “filing system.”

Besides all that, what sign do you have that people in the field of AI (or neuroscience) have any idea of what the human mind can do and how it can do it?

Simulating a human brain is a good idea that they an an inkling of that the human mind can do.

Basically all I have seen from this conversation and the Vinge essay is extrapolation from what already exists into an imaginary, science-fiction-like world.

An extrapolation of current trends is exactly what it is. That doesn’t make it any less valid. Yesterday’s science fiction is today’s reality.

[DavidV]

The latest issue of IEEE is dedicated to the Singularity.

It has a useful Who’s Who of all the big names in the pro and con camps. I’m not entirely convinced by the pro-Singularity position yet, but I’m certain that the positions of the featured opponents are bogus - they are Luddites and hippie philosophers.

[ruveyn_ben_yosef]

I’m wondering that after hundreds of years of trying to create an artifical intelligence just like a human being, scientists will throw back the curtain and show us a human baby born from a human being. :dough:

Pay no attention to that baby behind the curtain! — The Wizard of Oz

[volco]

I’m wondering that after hundreds of years of trying to create an artifical intelligence just like a human being, scientists will throw back the curtain and show us a human baby born from a human being. smile.gif

That’s a ripp off of Bradbury’s Fahrenheit 911

The latest issue of IEEE is dedicated to the Singularity.

It has a useful Who’s Who of all the big names in the pro and con camps. I’m not entirely convinced by the pro-Singularity position yet, but I’m certain that the positions of the featured opponents are bogus - they are Luddites and hippie philosophers.

How are you not convinced? I’m not defying you, I’m curious.

I’ve observed that ever since the industrial revolution, every “social movement”, be it socialism, nazism, or the Vatican, could only be defnied as Luddite (at least before I read A.R. and started calling them mystics).

The only problem would be extropy for extropy’s sake.

In this context, and I’m sorry that I can’t put this into better words yet, I think that Objectivism, a philosophy that retains human individuality, could be the ONLY way to profit and survive exponential technological growth.

Why?

Altruism would hold that, not now the planet, but the universe, is more important than the human race and its component individuals, and would hold extropy sacred in itw own “right”.

While Luddites often use technology in detriment of us all (they would be the million hands pushing the botton)

Objectivism stands that a lipstick is no more important IN ITSELF than a microscope, i.e. that worth is relative to the individual.

[Steve_D_Ippolito]

That’s a ripp off of Bradbury’s Fahrenheit 911

You either mean Michael Moore’s Fahrenheit 911, or Ray Bradbury’s Fahrenheit 451. If the latter, I don’t see how the singulary ties in to that *particular* bit of fiction, which was set in a world where there were no computers (mentioned anyway)–but that’s secondary to the point which was that there were no books.

[volco]

You either mean Michael Moore’s Fahrenheit 911, or Ray Bradbury’s Fahrenheit 451. If the latter, I don’t see how the singulary ties in to that *particular* bit of fiction, which was set in a world where there were no computers (mentioned anyway)–but that’s secondary to the point which was that there were no books.

What a horrible lapsus. I just pointed out the similarity of finding a real baby in a “too artificial” future, with finding a real book in a too digital future but I must confess I read that book way too long ago.

[ruveyn_ben_yosef]

What does this mean? My pocket calculator has more “power, speed, and utility” than me when it comes to arithmetic. Are you claiming that there is a special realm of computing restricted to biology and/or the human brain? What is the evidence for this?

Cardiac pacemaker. A computer controlled device for regulating arrhythmias of the heart. Also there is a computer controlled insulin pump that monitors blood sugar level and releases insulin when required in proper dosage. Not as good as the pancreas but it beats periodic injection of insulin by hand. I think there are now digital cochlear implants.

ruveyn

[israel]

What does this mean? My pocket calculator has more “power, speed, and utility” than me when it comes to arithmetic. Are you claiming that there is a special realm of computing restricted to biology and/or the human brain? What is the evidence for this?

weather or not we have the ability to, i think we should not create an a.i.

if we did, then we should not give it rights, and enable someway to destroy it at will.

think about it. if we were to create an artificially intelligent being, than not only would it be aware of its own existence, it would be aware of the follies of life.

it would understand that being that its consciousness is made up of nothing but electrical connections, it does not need to fear death. and as weve already started on technologies that allow machines to create themselves, it would never worry about decay.

In essence we would create a being that not only will never die mentally, but also never die physically (as it could infinitely rebuild itself and download its consciousness into new forms)

eventually it would come to the point where begin a machine it has no need for pleasures or pain, and will simply need something to do.

that thing to do?

take over the world.

[Jake]

weather or not we have the ability to, i think we should not create an a.i.

that thing to do? take over the world.

Oh no! Dun dun dun, du Dunh. Dun dun dun, du Dunh.

[Q.E.D]

You say that our computers cannot exceed the human mind? What happens when build computers OUT of human minds or out of biological neurons or what happens when we construct brains out of synthetic neurons with the same properties. To say that there is something magical about human consciousness is to speak in terms of mysticism and faith. I’m not sure if anyone here’s grasped the real issue concerning artificial intelligence, we’re OUTSIDE THE SYSTEMS which we’re discussing.

From inside our minds, we feel profoundly influenced by emotions, but those emotions are only electro-chemical signals being broadcast by different parts of our brains seeking positive feedback. On paper, we look just as ‘non-real’ as any sophisticated machine. The difference between our brains and the computer technology most of us are acquainted with is that our brains are both the hardware AND the software of human consciousness. No one opens us up to move around transistors <=> our brains aren’t solid state devices. However, if you’re interested in studying the human brain from a computer scientists perspective you might want to look into evolutionary programming algorithms.

AI is only a matter of time.

[CptSunbeam]

Sorry to drift off topic, but I’d like to say how ashamed I am of my response dated 26th December 2006, which is quite a while ago now. I’m actually a rather strict atheist now, having become one roughly around march 2007. At the time I was a typical literalist christian, and my arrogant comment reminds me of the sludge which I now have to deal with from others. I feel particularly bad about the emoticon I put at the end. In fact, my most fierce and arrogant online opponent over the past two years uses the same damned icon at the end of every post he makes, and I resent him for it.

Perhaps the lesson here is that no matter how intensely a person puts across their view, they may change their mind in the near future. Please excuse the diversion, enjoy your conversation. Actually I still love technology topics so I’ll enjoy it too.

[Jake_Ellison]

You say that our computers cannot exceed the human mind? What happens when build computers OUT of human minds or out of biological neurons or what happens when we construct brains out of synthetic neurons with the same properties.

I have no idea. Neither do you, in fact you have presented no evidence that it is even possible, let alone that it will happen.

To say that there is something magical about human consciousness is to speak in terms of mysticism and faith.

And hre you are, th first to say it in this thread (no one said consciousness is magical), and then double magically assign it to whoever you’re unnamed counterpart is in this argument, magically switching him out for a gentleman made of straw.

From inside our minds, we feel profoundly influenced by emotions, but those emotions are only electro-chemical signals being broadcast by different parts of our brains seeking positive feedback.

Please. That doesn’t even mean anything (the part about brainparts “seeking” things as if this were a fairy tale describing a magical journey, and the “feedback” was the elusive fair lady, is a damn metaphor you never even attempt to explain), let alone describe what emotions are. You have no ieda what emotions “only” are. The brain is nowhere near an open book we know everything there is to know about.

On paper, we look just as ‘non-real’ as any sophisticated machine.

What paper? You have not presented us with that elusive paper that contains the supposed mechanisms that explain every detail of my consciousness, because you don’t have it. It doesn’t exist.

The difference between our brains and the computer technology most of us are acquainted with is that our brains are both the hardware AND the software of human consciousness.

Bold statement, Why don’t you copy/paste the code of that “software” you know exists here? Or is this the first unexplainable software in human history?

No one opens us up to move around transistors <=> our brains aren’t solid state devices. However, if you’re interested in studying the human brain from a computer scientists perspective you might want to look into evolutionary programming algorithms.

Nothing you have said in this post relates in any way to evolutionary algorithms. I’m already aware of them, so just dropping the name, without even attempting to explain, leaves me to reply in similar fashion: Why don’t you check out Objectivism? It proves you wrong.

AI is only a matter of time.

That’s a prediction of a future that hasn’t happened yet. All predictions of the future are strictly the product of fantasy, not rational thought. I know for a fact that you have no idea whatsoever if AI will happen or not.

Now, if you would’ve said that AI is not impossible, that it might happen in the future, I would agree. I think I tried to explain why in this thread a while back.

[Jake_Ellison]

Sorry to drift off topic, but I’d like to say how ashamed I am of my response dated 26th December 2006, which is quite a while ago now. I’m actually a rather strict atheist now, having become one roughly around march 2007.

You should be proud of your achievement, not ashamed of your past, especially (but not only) if you became religious as a child.

This is easier to say than do, but the past you already corrected should not be what guides your emotions, but the latest relevant act (the change) should instead. And the change you’re describing is quite an achievement.

[Bastian_Hayek]

I became a huge fan of Ray Kurzweil and think his forecasts are based on facts. And I see a lot of Objectivists that think he is right. However, I was challenged and am not currently able to repudiate the claim by friend of mine on Facebook that following statement by Kurzweil is nothing but determinism.

I would say that it is not determinism but rather making forecasts on the basis of facts and patterns that are visible right now. What do you think?

“An analysis of the history of technology shows that technological change is exponential, contrary to the common-sense “intuitive linear” view. So we won’t experience 100 years of progress in the 21st century – it will be more like 20,000 years of progress (at today’s rate). The “returns,” such as chip speed and cost-effectiveness, also increase … Read Moreexponentially. There’s even exponential growth in the rate of exponential growth. Within a few decades, machine intelligence will surpass human intelligence, leading to The Singularity – technological change so rapid and profound it represents a rupture in the fabric of human history. The implications include the merger of biological and nonbiological intelligence, immortal software-based humans, and ultra-high levels of intelligence that expand outward in the universe at the speed of light.” - Ray Kurzweil

Determinism?