Banner
    Ray Kurzweil Pushes Singularity Back To 2045
    By Hank Campbell | February 10th 2011 11:58 AM | 134 comments | Print | E-mail | Track Comments
    About Hank

    I'm the founder of Science 2.0® and co-author of "Science Left Behind".

    A wise man once said Darwin had the greatest idea anyone...

    View Hank's Profile
    Last year I attended a singularity conference and Ray Kurzweil's avatar predicted it was 25 years away.   Well, it's been 25 years away for a long time.  It's a nice, safe number, close enough that no one gives up and stops buying books (global warming will happen in 100 years, for example) and not so close anyone looks silly (Al Gore saying in 2006 that we were doomed in 10 years, for example) it if doesn't happen.

     In 1993, for example, Vernor Vinge said "Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended." 

    Well, that would have been 2023.   In 2011 we can't even predict the weather 7 days out and brains are a lot more complex than weather so we will replace intelligence?   Not likely.   So Kurzweil later said it would be 2028 a while ago and now 2045.  

    In a recent TIME interview, Lev Grossman noted that Kurzweil was on I've Got a Secret in 1965.  Well, that's part of the problem.  He is stuck in the 1950s where virtually everything was believed to be distilled down to algorithms, logic, neural networks and maybe linguistics.   That, to Kurzweil, is what you need to reverse engineer the brain.   It's the same crappy 1950s thinking that led to lousy music quality we have today.   Who wants a brain that's analogous to CD recordings of real music?  Or, worse, MP3s?

    Kurzweil again falls back on the Magic Bullet of Future Science.   Like global warming skeptics who think nothing needs to be done today because "the future" will handle it, it's an intellectual placebo.   Because Moore's "Law" regarding computer processing (1950s thinking again) says computers will get faster, Kurzweil believes Faster Computers + [Magic Black Box Science] = Artificial Intelligence.

    It would be silly except he makes a lot of money at this.   Now, he is not in the same league as homeopathy, which bilks primarily poor, uneducated people.  Kurzweil is instead bilking people who are educated but maybe too optimistic, so if they aren't caveat emptor about his claims, we have no reason to deny him a livelihood.   

    And he still works in his old dates so he can't be called out too readily - he still says these magic exponential curves will successfully lead to reverse-engineering the human brain by the mid-2020s but the actual Singularity will happen by 2045.   The TIME scribe calls him conservative - well, he could be dead by then so it's conservative in the sense that it's a safe date.   Kurzweil's father died at 58 and Kurzweil is already 62 so living to 96 may be a stretch in his mind, though he claims his "biological age", whatever that means, is only 42.    So he can live forever if his biological age never changes.   Or something.

    The Singularity remains, for now, religion for technology geeks.   Like the guy on the street corner predicting the end of the world, they will eventually be right or die, there's no middle ground, but for people outside the movement it's hard to understand why anyone spends their days thinking about it.

    Kurzweil has always been an interesting person, a fun guy to sit around and talk with, if you can suspend disbelief.   But his belief that the biochemical complexity of an organic brain can be matched by silicon has to be maddening to biologists - especially when he falls back on the response that the complexity of reverse-engineering the human brain is not the problem, but rather biology underestimating the power of exponential growth.

    Comments

    Kurzweil's opinions represent one (near) extreme end of the Singularity spectrum. While many of Ray's specific predictions seem wildly 'out-there', Ray makes many more mundane predictions that, so far, have proven remarkable prescient. Those of us who follow the Singularity movement earnestly are well-read across MANY futurist publications - of which Ray Kurzweil is but one mote of light.

    It is increasingly difficult to discard the notion of 'smarter-than-human' constructs when we are presented with clear milestones on the way to that specific goal. IBM Watson on Jeopardy represents the ability to synthesize human language and complexity using brute force and Machine Learning. The creation of a Robot-wikipedia to share learned items across all robotics is another incubator for the future self-aware AI.

    Futurism and the Singularity is predicated on the sum of the whole of human scientific advancement, not on any one discipline, or one Futurists predictions.

    Kurzweil has always said the Singularity will happen in 2045. That has never changed. Check your facts Hank. Read TSIN. 2045 is in there as the date it will happen. He has NEVER said it will happen prior to that. There is nothing religious about it. He's Just observing empirical data and revealing it to people.

    Hank
    "Always" started in 2005.   In 2001 he didn't say 2045 and in the early 1980s, after he lifted the idea from books such as 'Two Faces of Tomorrow' he didn't say 2045.   Look, you can believe with as much fervor as you want.   Like with other people who believe in various forms of magical ascension, it's no matter to me.   You can also be gracious and say an iPod is a computer so people are wearing those computers he predicted we would all be wearing.   The great thing about fans of psychics, astrologers or anyone else is they will rationalize ways to make their prophets correct.
    Well hank its like this , Kurzweil has 19 patents 4 doctorates from 4 presidents and the 1998 medal of technology from the president of the usa . What have you accomplished hank? Hmmmm thats what i thought . I bet your a jesus lovin journalist itchin to prove your christianism. Well its like this , kurzweil continues to accurately predict technological achievements to this day. Hate on him all you want but he's right . He's a little nuts but he' s right. Now like i asked Hanky what have you accomplished ? Hmmmm?

    Gerhard Adam
    So your position is that arguing "from authority" is a way to gain credibility?
    Mundus vult decipi
    Hank
    But check this out
    I bet your a jesus lovin journalist itchin to prove your christianism.
    He sees into my brain like I am a research monkey!  Who had any idea anonymous sycophants on the Internet could diagnose people so well?  No wonder he believes in magic like the singularity.
    still I would rather listen to a man who has great creditability in his field then some nobody.

    Gerhard Adam
    Then by all means listen, because then you can all hold hands and worship when he's done talking.  After all, if you're that easily impressed by "fame" [since you are clearly unqualified to judge credibility], then you're a sucker to believe anything.
    Mundus vult decipi
    Ok, so you are asking me to believe you and not a man who is clearly much smarter and more creditable than the rest of us. I really don't know what you hope to accomplish here. It must suck though that if everything that Ray says is true, you will not embrace its benefits. If you do however, then you are going against everything you believe in. Its gonna must suck dying due to ideology. Even if I'm wrong life seems a whole lot better believe its half full then half empty. Keep on truckin my skeptic friend. I just don't see how you can effect anyone's minds about this subject. Quite frankly you are wasting your precious breath and time since your time in this world will be much shorter than the rest of us. I have time so I can afford to be here, you however don't. Stop being in the internet, stop trolling,go outside, and be with family. Life is literally to short for you to be here.

    Hank
    What is all this dying and ideology business? It's science, if he turns out to be right, he turns out to be right, there is no ideology.

    Unfortunately, Ray doesn't seem to believe in science.  He is falling back on magic.
    Gerhard Adam
    ...a man who is clearly much smarter and more creditable than the rest of us.
    ... and thus are followers found to carry on the new religion.

    You obviously have no clue about any of this, since you seem to think that it is akin to a religious experience, an ideology that one must embrace lest they be forbidden to "join the club".
    I have time so I can afford to be here, you however don't. Stop being in the internet, stop trolling,go outside, and be with family. Life is literally to short for you to be here.
    If you believe that then you're a fool..  If you think that YOU have time because Kurzweil says so ... well ... good luck with that. 

    I will make one prediction though ... you are heading for a very disappointing future.
    Mundus vult decipi
    I believe Kurzweil is around 75% correct in all his predictions. So if he's even 75% correct on the Singularity, we'll see it between 2040-2070.

    Hank
    Being X % correct does not work the way you think it does.  What it means is there is a 75% he will be correct with an error % of whatever - I would put his +/- at 75%, actually.  But it means there is a 25% he is totally wrong.

    If you said 99.7% you would be more wrong than he is, so kudos to you for keeping it real.
    Gerhard Adam
    I suppose it helps when even the slightest improvement is viewed as an overwhelming success.  I couldn't help but notice that Big Blue's one game win over Kasparov was reported as having "trounced" him.
    The machine is a grandkid to Deep Blue, the chess-playing IBM supercomputer that trounced world champion Garry Kasparov in 1997.
    http://money.cnn.com/2011/01/13/technology/ibm_jeopardy_watson/index.htm
    The simple truth is that none of these advances have been as easy as projected, nor as successful as claimed.  Originally it was supposed to be "expert systems" of 20 years ago that would revolutionize medical diagnosis to help doctors. 

    When it's all said and done, there will be decades worth of research and effort put into these projects, and even if they are wildly successful, they will end up being little more than super-fast Wikipedia computers.  I don't know which is worse, whether its the wild optimism of those hoping for such technologies, or the negative view they hold of human intelligence that they think its so easily replicated.
    Mundus vult decipi
    Bit of a silly article IMHO

    "it's hard to understand why anyone spends their days thinking about [the singularity]"

    It's a bit like someone in 1890 puzzeling why anyone would waste their time thinking about heavier than air transport or wireless communication or such nonsense that would never happen.

    It'll probably will happen and be the most significant historical event of the past million years or so, so I'm curious about it. You can always go back to you LP records and ignore all this newfangled digital data stuff that will never catch on if you like ;)

    Hank
    It's a good analogy.   Live music, or an analog physical recording with millions of timesteps,  is superior to 65,000 timesteps in digital CDs yet many people probably would not notice.   Millions of neuronal connections parsed down to 65,000 because Kurzweil believes that, since some are duplicated they must be redundant (which shows he was a good programmer in 1970 but still knows jack about science) means you will be left with an artificial intelligence is basically a more clever Cleverbot.

    But, hey, if you like MP3s and conversations with robots, more power to you.
    16 bit audio has much better resolution, much less noise and wider frequency response than analog tape. That was kind of a bad analogy.

    Although we are a far way off from AI as complex as our own, I would argue its almost "religious" to think that it could never happen. Our brains/minds are governed by physical laws which can be determined, and therefor can be simulated. I`m not gonna hold my breath, or dream of bringing dead loved ones back to life though.

    Hank
    16 bit audio has much better resolution, much less noise and wider frequency response than analog tape. 
    I never said anything about tape.  I said an analog physical recording.  If you disagree that an LP is closer to a live experience than a CD, you don't understand recording.
    FYI I have been a recording engineer for 5 years. Tape, the phonograph ect. are all analog mediums. Digital recording is far superior. It has lower signal to noise ratio ( 10-20 times lower) increased dynamic response, wider frequency response ( basically everything that makes audio higher quality).

    your observations are partially correct (although subjective). I agree that l.p`s sound more NATURAL, and therefor resemble live performances from pre 2000`s more closely, but that is because most analog recordings were mixed with very little compression and equalization. Now that digital effects are so common place, and listeners are so used to hearing music at high RMS levels, Live sound engineers are forced to use digital limiting ect. to compete with the sound of modern commercial albums.

    Regarding the subjective opinion that audio fidelity is getting worse, that is a result of the loudness war, not the switch from a analog to a digital medium. http://en.wikipedia.org/wiki/Loudness_war

    cheers!
    Liam Devlin

    Hank
    You're now making a different point.  I certainly agree about compression - one of the reasons a band like Toto still sounds likes its work was recorded yesterday was compression that other producers did not use at the time - but the goal of recording music historically was always to replicate the natural experience not everyone could enjoy in person.   Today that is different, of course.  I touched on this more fully in After The Singularity, Music Quality Will Be Even Crappier Than It Is Now but you are likely to disagree if you think .MP3s being worse fidelity than, well, everything, is a subjective opinion.   Physics is not subjective.
    Its not a different point. Digital recording is BETTER suited for accurately reproducing sounds than analog. More dynamic range, wider frequency response and significantly lower signal to noise ratio results in more detailed and accurate reproductions of audio. recording to tape (which is how l.p`s are recorded) actually compresses the audio, resulting in less detailed transient response. This is where you have things backwards. using compression results in LESS accurate reproduction of the source, because it is altering the dynamic range and frequency response of the audio, and bringing out harmonics. This is basic audio engineering stuff.

    The trend toward albums that sounding less "real" is not as a result of recording audio to a digital medium, it is caused by the power of digital editing and our tendency to favour louder music due to the psycho-physiological effect illustrated by fletcher munson curves.

    The one thing that we agree on i that physics is objective, which is why I have deterministic views of the world, but music and art is highly subjective. My point is that a bias towards "accurate" reproductions of audio is completely subjective and that a trend away from Natural replications of music is not caused by recording to a digital medium.

    sorry, how vinyl L.P`s WERE recorded in the 70`s, 80`s

    .mp3 is not the format that is used to record audio either. .aiff is the most popular format. It is used for all compact disk pressings. .mp3s are a very popular way of listening to music nowadays, but even .mp3 is higher resolution that analog audio (ei. vinyl, tape)

    Aitch
    As a former [analogue] sound engineer, I dispute that digital recording is superior
    The  actual dynamic range of music is so 'processed' that there is actually less dynamic difference between the lower parts and upper parts of the dynamic spectrum, due to the forcing of dynamics of even quiet passages, and this is because of broadcaster's preference for boosted dynamics, theoretically to make music sound louder on small radios and car systems
    Even the latest 'audiophile' digital-to-analogue-converters [DACs] still cannot attain the harmonic frequencies that 'old skool' class A and valve amps used to produce 30 odd years ago
    When have you heard a true ringing sound reproduced digitally? Try analogue!
    The so-called loudness war emanates, again, not from music, but from the advertising industry, in an attempt to get their product noticed, by making the sound accompanying adverts louder than the previous TV program material, and it spills over into 'production techniques'
    For what it's worth, the colour is boosted as well.....and as a practice, I have to say, it sucks, and most people hit the mute button or flip channels
    It's the main reason I don't have a TV anymore, well, that and BBC licence 'fees' [read tax]

    Fidelity, not just of music, is getting worse, because people just don't hear as much [or perform as much] live music in their own homes, so don't generally know what they are missing

    btw Kurzweil was a fool

    Aitch
    My point is that the trend towards decreased dynamic range has nothing to do with recording to a digital medium vs. recording to tape and everything to do with a social trend. You have said this yourself quote: "The so-called loudness war emanates, again, not from music, but from the advertising industry, in an attempt to get their product noticed, by making the sound accompanying adverts louder than the previous TV program material, and it spills over into 'production techniques'"

    The argument you are making is that analog saturation is preferable to digital saturation. (which is subjective opinion, many established AE`s debate this issue) In my opinion modern tape saturation plugins are able to reproduced the desired tonal coloration, and are easier to control digitally, because more parameters can be altered. This argument (like hanks) is subjective opinion, and has no relevance as a rebuttal to my argument that digital recording mediums are able to reproduce sound MORE ACCURATELY than analog tape (which is the medium you and Hanks favourite albums were recorded to)

    The fact of the matter is that 16 bit .aiff, .mp3 and.wav is MUCH higher resolution than tape, or any analog medium. Because digital recording results in a more accurate replication of sound (doesnt compress the audio, resulting in increased even order harmonics), it is less pleasing to the ear. This is NOT a subjective opinion. this is FACT, and anyone who is a serious audio engineer will understand this.

    Recording digitally does not prevent you from using analog gear to add analog compression and saturation to your audio. Because digital conversion is so transparent it is possible to send your digital tracks out of the computer, run them through an analog compressor, and then route them back into protools, which allows you to reap the benefits of analog coloration without the degradation of recording to tape. Also, most professional producers/engineers will track through a high quality analog mic pre and dynamics processor before going into protools which gives analog coloration.

    Modern digital to analog conversion technology is very detailed and accurate. The degradation of an audio file that has been passed through a digital to analog conversion stage 3 times is virtually indistinguishable from the original file in a double blind test. To reach the level of degradation of analog tape, digital audio has to be passed through 20+ stages of digital to analog conversion. (this can be heard for yourself at 43:30 of this vid http://www.youtube.com/watch?v=BYTlN6wjcvQ I suggest watching the whole video, it will explain in greater detail all of the points I have been arguing. if you really think the youtube sound quality makes a big difference, download the .wav files from ethan winers website and check them out!)

    Here is another video to support my argument:

    http://www.youtube.com/watch?v=AgH99wyRoVc

    cheers
    Liam Devlin

    And its funny that you guys are being so derogatory to me. It is clear that you are very ill informed when it comes to recording audio, or even understanding the debate of digital vs. analog. Im not here to argue which is better, or to argue for or against ray kurzweil, I`m merely pointing out a very bad analogy in Hanks article. Maybe he should learn more about audio before using it as an analogy in his articles. I`m not arguing about whether analog saturation is preferable to clean digital recording, I`m arguing (as pertains to hanks analogy) that digital mediums are capable of reproducing sound MORE ACCURATELY than analog mediums.

    Hank
    And its funny that you guys are being so derogatory to me.
    It's not derogatory to note you keep changing your argument and were misrepresenting mine.   You decided my argument was about friggin' cassette tapes despite my not saying anything of the kind and declared victory noting the fidelity of those was awful.  It was a straw man.  Then you asserted .mp3s were as good as LPs and then changed your argument again.
    Im not changing any argument. .mp3`s/.wav/.aiff/ any digital medium are more capable of accurately reproducing audio than analog tape/vinyl. Im not talking about cassette tapes, Im talking about 8" reel to reel that was used to record all vinyl records pre 1990`s. FYI It is possible to record digitally, then press the digital mix to vinyl!

    Your original argument was that digital mediums are not as capable of ACCURATELY REPLICATING SOUND as analog formats such as tape, vinyl ect which couldnt be further from the truth. Digital Recording via .mp3, .wav or .aiff is as pristine of a medium for capturing audio as there is!

    This is a bit like trying to convince a creationist of evolution. watch the videos I posted and consider for a second that maybe you are mistaken!

    In your article you stated that the reason vinyl/analog recording sound better than digital is because digital information is buffered. But the negative effect of buffering the audio is more than compensated by the better resolution, increased dynamic range, lower signal to noise ratio, decreased susceptibility to environmental changes ect. ect. ect.

    you have mistaken a subjective argument for a factual argument concerning the fidelity of digital audio

    Aitch
    Actually, you originally said,
    "Digital recording is far superior. It has lower signal to noise ratio ( 10-20 times lower) increased dynamic response, wider frequency response ( basically everything that makes audio higher quality)."
    I disagree, surprisingly, as Hank stated in his other thread - one is continuous, the other samples (which is nearly everything that makes the difference to higher quality!)

    You are mistaking 'accurate' for 'musical' and 'aesthetically pleasing' which sampled sounds still don't achieve, not even at 24 bit/192 khz sample rate

    Far from being ill-informed about audio, I am extremely well informed about my chosen field, analogue audio, but have also compared the best digital equipment, and I still prefer analogue amplification, even if playing digital sampled music [CD/ your Youtube samples through it]
    btw in your 1st youtube link, that is not the whole picture, as slew rate in an audio amplifier makes a huge difference, yet is not mentioned at all, although they, too, use analogue amplifiers to listen to the final results of comparisons! Also, it is false testing to limit audio frequencies to 20-20khz, as harmonics far exceed these frequencies on many stringed instruments, brass, woodwind, piano, and drums..... and that is digital's main failing in my mind

    I think, again I understand Hank's point relating the deteriorization of digital sounds, to the deteriorization of the digital information proposed by Kurzweil as 'the singularity' at a point in the future - when 'superior' digital bot replaces natural human genome'

    Given the complexity of the human genome to a 'simple' natural audio sound, I think Kurzweil is way off, and your notion of audio accuracy, seems to fall into the same delusion

    I think Hank has greater appreciation, as a musician, of the analogy used, than you do, and appear to still be learning about digital recording techniques

    Moreover, I think you misunderstand the concepts of what makes natural audio preferable, and more pleasing to the soul, than digital

    Had you included a few more analogue amplification and recording articles into your comparison, you might not feel that the audio debate is so cut and dried......and that actually it is closer to the dilemma about 'future artificial intelligence' than many people realise

    Aitch
    Gerhard Adam
    Our brains/minds are governed by physical laws which can be determined...
    Why is it that our brains are governed by deterministic physical laws when it comes to artificial intelligence and they're indeterministic when it comes to "free will" arguments. 
    Mundus vult decipi
    Thats because free will isnt indeterministic, which is hard for most to grasp, because it goes against our intuition much like Special and General Relativity. I was once a proponent of free will, but after reading Sam Harris`s work, and learning more about evolution, physics and neuroscience I think otherwise.

    Gerhard Adam
    Trust me... I'm no proponent of "free will".
    Mundus vult decipi
    SynapticNulship
    There's no free will, but I can sell you some real cheap.
    Well I just checked out Cleverbot which I had not heard of before and have to admit it lacks a certain something conversationally. Even so computer progress seems to be cracking along what with chess, jeopardy, Googles self driving cars and the like. Not that it's necessarily good or bad but it seems to be happening.

    Gerhard Adam
    Certainly some progress in computer technology is being made, but when one considers that Deep Blue (vs Kasparov) only won by one game after which 14 years later we have a 2880-core Power7 system that will play Jeopardy, it doesn't seem like it's been that significant.

    In perspective,  these problems always seem to get "solved" by brute force which suggests there's been little or no movement on actually understanding what intelligence is, let alone figuring out how it can be used effectively.  Even the Watson system for Jeopardy seems like little more than a huge search engine that works only because of the massively parallel processes involved.  After all, if you can effectively search millions of documents and prior Jeopardy questions in a short time, you may win, but I'm not convinced there's any "intelligence" involved.

    It is a bit annoying when you see some reports that suggest the two Jeopardy opponents are the "best" that humanity has to offer. but that's obviously media hype.  What would have been nice would be to see how well Watson would've fared against Kim Peek (the "real" Rain Man).  I would've paid to see that.
    Mundus vult decipi
    Hank
    Yes and no.   In numerical modeling, for example, much more powerful computers mean simulations can be done faster.   It does not mean they will be more accurate.  Likewise, better computers that can parse words fast, store more data and analyze more pieces in a game with finite rules does not mean a replacement for the human brain is any closer.   They are apples and oranges.

    Kurzweil is hoping for the biggest 'black box' in human history to happen between us and true AI and make him right.   And that's all it is, hope.  No actual advancements have been made in algorithms in 20 years.   Faith is fine, of course, billions of religious people believe in an ascension as well.   But stick some money in your 401K just in case. 
    While your argument is definitely well thought out, I'm not so sure I would have used weather forecasting as a comparison. Having used many numerical forecast models operationally and from a research perspective, I can personally tell you that we've made leaps and bounds in this area. The physics and parametrization packages we use are becoming exceptionally robust. The big hamstring in our arena is the quality of the data ingest. 70% of the world is open ocean, so we have to rely on less accurate satellite based soundings and lots of interpolation. This is something other sciences likely do not have to deal with (or nearly so much). A 6 day forecast now is about as accurate as a 3 day forecast 15-20 years ago, which is amazing considering the data ingest hasn't improved all that much.

    Hank
    Chris, you certainly have a point; I could have used an analogy of solving for electromagnetics issues using Maxwell's Equations but it would have resonated with a lot fewer people.   Weather is complex, and chaotic, and forecasting such requires multiple variables so, like Maxwell's Equations, can never really be solved, but instead better and better answers will instead be converged on.

    Our brains are likewise more than a clump neuronal connections which can be mapped on a piece of silicon and therefore be considered to mimic human thought.
    On that same vein, without a good working knowledge of the brain's inner workings (good data), our (future) approximations or imitations of human intelligence aren't going to be very good. Some of Ray's ideas certainly seem pretty plausible, but if I've learned anything in the weather field, it's that things never work out quite the way you think they will. Trying to forecast the future in any capacity is often a humbling experience.

    Dave Whitlock
    The singularity will never happen. “Intelligence” is not something that can be designed to occur at an arbitrary level. I don't doubt that machine substrates can support intelligence, but they would need to be designed to exhibit the degree of intelligence that they exhibit. The problem with the singularity is that there is no way for a machine of a given intelligence to design a machine that is more intelligent than it is. How can an entity of lower intelligence identify which of two high intelligence entities is actually intelligent and not simply mad? Determining the intelligence of another entity requires pattern recognition. If the only intelligence pattern you have is your own, how do you tell if a pattern that you are unable to understand is of a superior intelligence of or an insane intelligence?
    "There is no way for a machine of a given intelligence to design a machine that is more intelligent than it is".

    Ahem...Suppose that some humans design a virus that reproduces itself quickly, that kills humans, and that humans themselves can't stop or cure.
    Would you say that the virus is less intelligent than humans?
    That would qualify as a singularity to me.
    Perhaps the virus would end up calling this the big bang in a few million years...

    Gerhard Adam
    ...and what does that have to do with anything.  You could apply the same argument to a nuclear exchange.  So what?

    Mundus vult decipi
    Dave Whitlock
    If you want to change the definition of the term "intelligence" to something a virus can express, we are not talking the same language. If you want to change the definition of the term "design" to also include a few million years of randomization and evolution, they we are not talking the same language.
    Look at a chess player.
    He has intelligence, a set of rules, and a goal (winning).
    Now look at how a chess player learns how to play:
    Find patterns. Experiment moves. Find meta patterns. Experiment. Find meta-meta patterns...
    I see a similarity with a virus that mutates. The time and space scales are different.
    I find the virus more perplexing because it is so much simpler than a human.

    Regarding design, I don't have the beginning of a freaking clue. I'm opened to enlightenment
    If you can get everybody else to agree on it, that is :-).

    Gerhard Adam
    Except that the virus isn't directing its mutations.
    Mundus vult decipi
    Well...if he virus mutations are due to random events like some atom or molecule interfering with the process in very unpredictable ways, I still find an analogy with the possibility that the player finds a new "experiment" subconsciously.
    Subconsciously meaning something random like the thought of a castle in last night's movie pops up as a move of a tower in the game, just because of "brain noise".
    I think that the difficulty here is that the brain has difficulty working with large numbers, such as the number of neurons in a brain, or the number of viruses that are allowed to die without having produced anything useful.

    Then again. Thinking about your own brain is tiring.
    Plus storing information that's already there is kind of stupid :-)

    Hank obviously hasn't read Kurzweil's books or he would have noticed that most of the points that he brought up have already been addressed. And the singularity isn't defined (by Kurzweil) as the year humans become immortal... it's the year that machine intelligence becomes better at improving itself than humans can. In fact, after just a few paragraphs I'm wondering why Hank bothered to write on something that Hank knows nothing about.

    ...and he doesn't believe in global warming.... hmmmm...........*rolls eyes... ok then. That's a little bit of a credibility killer. I'll bet you go to church every Sunday too lolz. Just the kind of person that I'm to get science info from.

    Gerhard Adam
    ...it's the year that machine intelligence becomes better at improving itself than humans can.
    A fantasy by any other name is still a fantasy.
    Mundus vult decipi
    I've seen many compelling arguments supporting Kurzweil, and many that suggest that's his ideas are ridiculous. Hank's article is just ridiculous to read, does a horrible job of addressing the questions and doesn't even build his own credibility by actually addressing what Kurzweil has already said.

    And while I personally don't agree with most of Kurzweil's claims, computers being able to improve themselves faster than humans can is almost certain to happen:

    Do some quick wiki reading on genetic programming if you want to see what i'm talking about.

    Gerhard Adam
    ...computers being able to improve themselves faster than humans can is almost certain to happen:
    Not in the way being envisioned.  Certainly there are algorithms that can be applied to improving how systems behave, but they will always simply be less than "intelligent" in any such sense.

    I won't repeat all the arguments, but essentially any machine that possesses a human like intelligence is simply a simulation.  Unless someone is willing to provide a coherent definition of what "intelligence" is that doesn't associate it with problem solving for the organism, then any attempt to make a machine that behaves like a human is simply a parody.

    Secondly, as has been stated before, you cannot produce an intelligence greater than your own (including a machine).  The problem is intractable since nothing produced by the more "intelligent" machine can be verified by a lesser intellect.
    Mundus vult decipi
    So by ur argument, complexity always requires something more complexity to create it and therefore complexity can never increase... hasn't evolution (and several other thought problems) proven this incorrect?

    Where is the evidence to support your theory that *anything generated by a machine is simply a simulation? If it thinks like a consciousness, generates the same behavior as consciousness, etc, etc, etc then it is for all intents and purposes, conscious. To say otherwise would be like arguing that other people in your lives are not really conscious... only simulations because you aren't experiencing their thoughts. What other way can we define intelligence that to describe it as a behavior?

    If you believe that consciousness is a nature occurrence... meaning it is a natural phenomena as opposed to super natural or magical, than you have to agree that the chemical, electrical signals in our minds are natural (globally) passive processes that are occurring naturally and following the laws of physics... logical laws of physics, and logic can always be replicated in a machine given the right software and powerful enough hardware.

    Gerhard Adam
    So by ur argument, complexity always requires something more complexity to create it and therefore complexity can never increase... hasn't evolution (and several other thought problems) proven this incorrect?
    This has nothing to do with it, since you're not talking about machines "evolving", you're talking about building them.  More specifically you're talking about building them and using them to your own purposes.  Therefore, you cannot build a machine more intelligent than yourself, since you'd be incapable of confirming you'd achieved it.
    If it thinks like a consciousness, generates the same behavior as consciousness, etc, etc, etc then it is for all intents and purposes, conscious.
    Not at all.  First you have to define what you mean by intelligence and consciousness, but to date, everything has an intelligence and consciousness commensurate with its requirements as a biological organism.  Ants don't think like bears, and bears don't think like humans.  We would clearly argue that such an arrangement makes no sense.  Similarly, if you built a machine that thought like humans, it would be a simulation since it isn't human and therefore a significant part of what humans require and utilize as intelligence simply isn't applicable.  If a machine were to ask me to join it for dinner, it would be ridiculous, since a machine has no such need.  So to produce machine intelligence, you'd have to have an intelligence that is oriented towards the "needs" of the machine without regard for humans.  In that respect, all of our efforts have been towards producing human intelligence simulators and not machine intelligence.

    In your last paragraph you shift from intelligence to consciousness, and yet despite however "natural" it is for biological organisms, it cannot be argued that it is "natural" for machines.  Therefore if you're simply attempting to duplicate something you find in nature, you've simply simulated it, since there was no evolutionary path to its creation.  Despite some arguments claiming that this represents a privileged status to biology, the truth is that this is a privileged position.  Cars don't have a need to travel without humans directing them.  Computers don't solved problems without programmers coding the rules and humans applying those rules.  In short, a consciousness implies a separate motivation and position with respect to everything else, and consequently a machine cannot simply be "conscious" unless it also possesses those elements.  Otherwise, as in the previous case, it is simply emulating what humans would do.

    Will your proposed machine have the ability to voluntarily cooperate or not?  Will it be able to deceive?  Will it be able to choose it's own activities?  If not, then it's still just a machine and, at best, is simulating the behavior we would like it to portray.  If it is truly independent and capable of true "consciousness", then we will have created a new species and will have no rights in determining what it does. 

    Asimov's three rules of robotics, presumes that robots aren't truly intelligent and will remain subservient to humans.  If these truly represented individual conscious intelligences, then such a situation would be ethically untenable since we will have simply rationalized enslaving an intelligent species to our own ends. 

    In the end, people are confusing the mechanics of intelligent behavior with actual intelligence.  What motivates a biological organism is its need to survive and the decisions it makes, as well as the lessons it learns are all based on the "knowledge" that mistakes lead to death and extinction.  Note I'm not proposing that every biological organisms as the cognitive ability to analyze this, but rather that it is an emergent property of simply being a biological organism.

    Unless you're proposing that biological organism (including humans) are simply algorithm machines, then you'll have to explain why you think a machine intelligence would be comparable to a biological one.  After all, would that include the idea that calculators "know" math?  Do word processors "know" how to produce documents?  If not, then what distinguishes "intelligence" from the mere mechanics of intelligence.
    Mundus vult decipi
    Dave Whitlock
    Continuity of consciousness is an illusion. It is an artifact of the computational weaknesses of the entity detection pattern recognition that humans have. It is analogous to optical illusions being due to idiosyncratic weaknesses in human visual detection and decoding computations. The entity of today is not self-identical with the entity of yesterday, or last year, or 20 years ago. The “entity” is at those different times is different than the entity at all other times. Humans are unable to perceive that they are not self-identical to an earlier self because their self-identity pattern recognition doesn't have the computational resolution. Why should it? There is no evolutionary advantage to a self-identity detection module with a resolution better than is needed to highly prioritize self-survival and self-interest. I could program a computer to “think” it was anything, including that it was Kurzweil. Just have it respond to every question with “I am Kurzweil”. Pretty clearly if every question is answered with “I am Kurzweil”, there will be no evidence that the machine does not think that it actually is Kurzweil. I appreciate that such an assertion would not be satisfactory, but there is no assertion that can be made that would be satisfactory. Any assertion is only the self-proclaimed assertion that “I am Kurzweil”. That is the problem that humans have. Their self-identity module simply proclaims “I am me” when interrogated. That is an easy thing to do. I can even do it with analog, or with just a cardboard sign. No one has any data with respect to the internal conceptual representation of what ever entity is making that assertion, even if the “entity” is a cardboard box with blinking lights. When people have serious brain injury, they don't start thinking that they have become someone else. Their self-identity module still proclaims when interrogated that “I am me”. When brain damage can seriously degrade all other cognitive, sensory, memory, and computational functions, yet the “I am me” conviction remains as strong as ever, this demonstrates that the “I am me” determination doesn't require a whole lot of computational resources. If it doesn't require that many resources, it can't be doing a high fidelity computation.
    Aitch
    Continuity of consciousness is an illusion.........
    I do believe you are confusing consciousness with ego
    It is an artifact of the computational weaknesses of the entity detection pattern recognition that humans have
    Computational weaknesses?....

    Speak for yourself!

    Human consciousness is far more than a mere computational entity, which is why machines cannot replace it

    Well, that, and the egos that get in the way....  ;-)

    Aitch
    Dave Whitlock
    Can you distinguish between the "you" of today and the "you" of 10 years ago? Which "you" is the real one? "You" have been under constant change since "you" were in utero. Which one of those is the "conscious" one? You can't distinguish between them because you don't have the data in storage to represent the earlier "yous" and you don't have the computational resources to compare all those different versions as represented by data.
    Aitch
    I'm not a computer
    My 'you' i.e. 'me' is the same as it's always been in essence, and is as conscious as it is, which is a continuum beyond my ego, my thinking mind, that greedy thing atop my body residing in my head....but you won't find 'me' in there....I flow between heart and head and belly and sex and senses, and gorge myself on life in all its forms of physical, mental, and spiritual delights

    Have you tried using an antivirus, or getting updates? ;-)

    Aitch
    Dave Whitlock
    Are you saying there is some non-material "soul", or other non-material entity that is the "real" you? Do you have any data to support your belief? Any data to suggest how a non-material mind could couple to the CNS and ANS and cause macroscopic effects and not violate conservation of mass, momentum, charge and spin? Any data beyond self-reports of people that they feel this way?
    Aitch
    My certainty trumps your doubt, and I'm as connected to reality as to be relating to you within it, so if I'm not here, you're talking to yourself, and I should see a doctor if I were you...

    That's all the data I need ...

    Get your own! If you can figure out how to look for it....

    The problem you seem to have is, that you think there's a disconnect between the soul, i.e. the real me, and the physical, ...well OK, turn it on its head,

    What evidence do you have that there's no soul in your physical body, and I'm talking just to your 'computing capability'?

    I don't think you can show I'm violating any laws of physics, either....


    Aitch
    Dave Whitlock
    There is not a single datum (note singular form) that is inconsistent with the non-existence of a non-material entity called the soul. The idea of the soul was not made up to explain some observation or data, it was made up to soothe the feelings of people (such as yourself) who don't like the idea that there isn't such a thing. If you want to believe in something for which there is no evidence, which makes no predictions and which cannot be analyzed, go right ahead. But you are not doing science when you do that. From information theory, it is known that changing information requires dissipation of energy. Is your “soul” actually doing anything? If it is, where is the energy coming from? Where is it going? If there is a soul that does all the tricky stuff that humans think up, why do humans need such a big brain? Why does messing up the brain always mess up the ability to think? You can speak about your certainty, YECs are certain the Earth is only 6,000 years old. Do you have some data?
    Aitch
    There is not a single datum (note singular form) that is inconsistent with the non-existence of a non-material entity called the soul.
    Ooh! I do love double negatives.......How many double negatives make a positive?

    You're just spouting someone else's words, not your own self-realisations......How do you know who YOU are, in that computational errorzone you call a head?

    What do you know of SOUL if you haven't got one....?

    ...and what my soul does requires mutual exchange of energy - if I don't give it any attention, it doesn't give me any guidance, its a co-dependent relationship, which I'm happy with, and your not knowing of it doesn't affect my relations with it at all, and if you say it's not science, where's your evidence other than saying it doesn't exist/you can't find it/don't know where to look?

    My certainty about my inter-relationship with my soul, and what YECs believe is as far from me as the 6000 years they believe their world has existed for

    My connection goes back 16+billion years...see the Ian Lungold video's referenced here

    http://www.science20.com/aitchs_hangout_come_and_join_me_wont_you/blog/e...

    and yours?

    Aitch
    SynapticNulship
    There are many forms of externalism which do not resort to old-fashioned dualism.  For instance, some externalists like Alva Noe considers consciousness to be the interactions of a dynamic system--a CNS interacting in a body interacting in a world.
    Gerhard Adam
    Isn't that just a convoluted way of calling it an emergent property?
    Mundus vult decipi
    SynapticNulship
    No.  To also be an emergent property requires, if I am using that term correctly, that the interactions making it up are simpler than the whole.  I think that is a differently claim.  Also, simply saying it's emergence doesn't specify where.  Scientists like Christof Koch are of the philosophy that consciousness is completely within the brain, whereas some externalists would say you need to include more of the environment or other types of interactions.  Both of those views could still be looking at emergence.
    Gerhard Adam
    Personally I think most of it is nonsense until we can arrive at an acceptable definition of what we even mean by "consciousness".  This is the problem with most of these kinds of discussions, because without a definition virtually any speculative concept might be considered valid.

    After all, does consciousness require "self awareness"?  What do we mean by "self awareness"?  Are there criteria that have to be met?  Can one be conscious without being self-aware?  What does it mean to be aware of unconscious processes or events?  Does consciousness require "control"? 

    I realize too many of these terms are biased around human perceptions, but without any reasonable definition, no progress can possibly be made.
    Mundus vult decipi
    Bonny Bonobo alias Brat
    Isn't 'the Singularity' the technological creation of smarter-than-human intelligence? I don't see why a smarter than human intelligence necessarily has to be necessarily conscious, surely it just has to be smarter and more intelligent? Artificial Intelligence, direct brain-computer interfaces, biological augmentation of the brain, genetic engineering, ultra-high-resolution scans of the brain followed by computer emulation are all different technologies which, if they reached a threshold level of sophistication, would enable the creation of smarter-than-human intelligence, however surely none of these have to be conscious in order to achieve this, couldn't they be smart and unconscious? Like people talking in their sleep but still making sense and in this case giving brilliant answers to any questions put to them.
    My latest forum article 'Australian Researchers Discover Potential Blue Green Algae Cause & Treatment of Motor Neuron Disease (MND)&(ALS)' Parkinsons's and Alzheimer's can be found at http://www.science20.com/forums/medicine
    SynapticNulship
    Helen, I agree that no machine needs to be more conscious to achieve anything.  Indeed, I'm not sure what more conscious means.  Did somebody say that in this thread?

    Isn't 'the Singularity' the technological creation of smarter-than-human intelligence?
    Actually no.  It's called a Singularity because that is a metaphor for the concept of a change so radical that there is no turning back and no way to predict the future before that event.

    So, that means all of humanity or some large subset has to be radically changed and/or become that smarter-than-human intelligence.

    Many people on the internet talk about "the AI" and that once a single smarter-than-human general purpose intelligent machine is built, suddenly at that moment the shit hits the fan and humanity is never the same again.

    But in reality, unlike science fiction, making one instance of a technology usually accomplishes nothing.  Going from one-offs to canonical widespread techniques may be accelerating, but even if it is, there's an infinite number of types of AI that can be made so it doesn't even make sense to say that there is The One (see how it starts to sound like a religious prophecy, or a movie like The Matrix?)
    Bonny Bonobo alias Brat
    Thanks Samuel, but now I'm even more confused (so what's new?).
    Actually no.  It's called a Singularity because that is a metaphor for the concept of a change so radical that there is no turning back and no way to predict the future before that event.
    Like the singularity in the center of a black hole you mean? Surely then there are many singularities in our lives, like giving birth for example? Is there a definition for the singularity that this article is discussing anywhere?
    My latest forum article 'Australian Researchers Discover Potential Blue Green Algae Cause & Treatment of Motor Neuron Disease (MND)&(ALS)' Parkinsons's and Alzheimer's can be found at http://www.science20.com/forums/medicine
    SynapticNulship
    Actually, I don't think you're confused.  You are right, you could consider lots of events in your life to be singularities. 

    Kurzweil's book The Singularity is Near doesn't give a concise definition--the closest he gets is a bullet point list of 37 items followed by the sentence "This, then, is the Singularity" (pp. 25-29).

    Bonny Bonobo alias Brat
    Thanks Samuel, I'll get it from the library this week. By the way I'd edited out 'more conscious' as it was a typo, so no one had said it anywhere.
    My latest forum article 'Australian Researchers Discover Potential Blue Green Algae Cause & Treatment of Motor Neuron Disease (MND)&(ALS)' Parkinsons's and Alzheimer's can be found at http://www.science20.com/forums/medicine
    Aitch
    If you believe that consciousness is a nature occurrence... meaning it is a natural phenomena as opposed to super natural or magical, than you have to agree that the chemical, electrical signals in our minds are natural (globally) passive processes that are occurring naturally and following the laws of physics... logical laws of physics, and logic can always be replicated in a machine given the right software and powerful enough hardware.
    I go along with the first bit...consciousness is natural, but find the 2nd follow-on a non sequitur
    I don't see consciousness 'following' the laws of physics, but maybe visa versa is true
    The problem with trying to make machines 'intelligent' is completely different to making them 'conscious' and logic has little to to with it

    Consciousness is not about logic - you should get out of your head more often  ;-)

    Aitch
    Aitch
    There's a whole heap of difference between artificial, machine 'super-logic', and Human Intelligence

    I don't see any of the current thinking on programming giving rise to intelligence....though it will undoubtedly get used for data analysis of human habits...so stupidity should soon get identified ;-)

    Aitch
    We already know how FM-2030's predictions for that mysterious, far-future year 2010 turned out:

    http://www.scribd.com/doc/10948503/Up-Wing-Priorities

    Also good for some laughs:

    http://www.fm2030.com/

    Hank
    I, for one, look forward to our transliving, hyperfluidian, Kurzweilian overlords of 2010.   

    His mistake was only thinking 29 years into the future.  If, like Kurzweil, he had made it an even 40 he would still have time to be correct.   And then he could blame a lack of 'exponential' thinking when it didn't happen.
    When one speaks in the name of Mayan prophecy then one is speaking in the name of Ray Kurzweil. Thus men will experience a leap in knowledge epodentially greater than modern times. When is not the main point but relative to the act. The act or event is strange to speak of but the mind can conceive of. He is just a modern perspective of things to come...

    Aitch
    When one speaks in the name of Mayan prophecy then one is speaking in the name of Ray Kurzweil.

    er, no!

    Aitch
    Every religious philosophy predict a singularity 'New heaven and new Earth forever and forever', ascension of the soul or spirit, etc. To think intelligent design e.g. artificial intelligence is a journey in to a limited vessel called technology. Timing is not the determining essence of the journey. Dates are measurements of fear and selfishness. Preparation is the message of all ages. These segmented commits are but points to connect in the grandest game of life. Thanks for all your efforts, thoughts and dreams.

    SynapticNulship
    Kurzweil has always been an interesting person, a fun guy to sit around and talk with, if you can suspend disbelief.

    Unfortunately, the one time I talked to him in person (about 5 years ago), he seemed to be under the weather and surprisingly uninterested in students building robots.  I came away very disapointed....he was supposed to be biologically younger than his true years and yet in person he seemed like a 70 year old about to fall asleep.  In his defense, however, at various presentations I've seen, he was much more energetic.
    I read the article in this weeks Time magazine and I think the dude is a fan of the New Version of Battlestar Galactica and Caprica. If there will be Cylons by 2045 then I guess I better be nicer to my toaster and my Norelco razor and my HP Windows 7 computer now. Lol if that is the case then I hope they make a Number 6/ Tricia Helfer Model like on Battlestar Galactica. Lol

    Hi, i'm pretty certain you are wrong about the 2028 date. Kurzweil made a bet back in 2003 that a computer would pass the Turing test by 2029 and has maintained that superintelligence would arrive some years after any Turing test was passed (he's predicting at least a 16 year gap if the singularity is supposed to happen around 2045) so a prediction for a singularity by 2028 doesn't line up with this.

    As far as i can tell, the only way you could be correct that he said 2028 for a singularity would be if he said this before 2003 when he made the Turing test bet and presumably it would have been quite some time before that. I think if you are going to write an article which is essentially bashing someone you need to get these things right, although i'm not saying you were deliberately misleading, to be clear.

    Gerhard Adam
    Doesn't really matter, but it is amusing regardless of when it is supposed to happen.
    Mundus vult decipi
    omg... Gerhard. You are in for a big surprise. I read all your posts. You're wrong. Regardless of what Ray says...it's coming. I knew it well before I read his book. You're the one in the dark.

    It's not even worth arguing. Wake Up. The Singularity is closer than you think.

    Look....I don't thin anyone really knows....this is all conjecture....but will something like the singularity happen? Most certainly. It may seem impossible to replicate the human brain....but it is much simpler than it might seem....you have the same respect form the brain that fellow crhistians once had for religious observance...

    You'll see....it's closer than you think, kurzweils timeframe is irrelevant. The more processing you have the more likely a discoveryvwill benmade thatcwill yield to new insights...once a a basic puzzle piece is found....it will be pretty fast from there...it's self reenforcing, the more technology, the more capacity to process experiments and discover, and the more discovery, the more processing power...you can call it whatever you wish.

    Btw, you sound like one stupendous ass.

    Its always good tovhave criticism though, you don't want tovget bilked by a fanatic...butntrust me, ithe brain is much simpler than you think.

    Gerhard Adam
    Famous last words ....

    I'll definitely take that to heart, especially coming from someone that understands neither computers nor neuroscience.
    Mundus vult decipi
    Hello. I am doing a little research about Brain computer Interface. Brain Computer Interface technology provides a direct electronic interface and can convey messages and commands directly from the human brain to a computer. It involves monitoring conscious brain electrical activity via electroencephalogram (EEG) signals and detecting characteristics of EEG patterns via digital signal processing algorithms that the user generates to communicate. So.. Yesterday I found one great Open Access ( free to download&share ) book “Recent Advances in Brain-Computer Interface Systems” here : http://www.intechopen.com/books/show/title/recent-advances-in-brain-comp... Take a look, it is great. Cheers! ;)

    Ray has been predicting the Singularity to occur around 2045 for quite a while. He has predicted that we would have computing power equivalent and then greater to a human brain much earlier than this, but that the software may not be sophisticated enough to properly make use of it until around 2045.

    Wow! The ignorance of this article is downright ballsy! You do realize that any prediction of the future is done blindly, right? So the very problem you identify as underlying Ray's argument also underlies your own. But you're actually arguing AGAINST technology and the nature of humanity to always grow and expand. That's about as Luddite as a point of view gets. I love that you posted this. It will provide great insight into the ignorance of our era.

    Since you clearly haven't read TSIN (or at least haven't read it closely), here's your basic misunderstanding: Ray's "predictions" are nothing more than extrapolations based on current trends and trajectories. Everything boils down to a simple identification he noticed, which he calls the Law of Accelerating Returns, which says that information technology evolves at an exponential rate, rather than a linear rate. That's it. That's the whole basis for his ideas. And his book presents chart after chart demonstrating the dynamic. If you want to challenge his ideas, you have to discredit those charts. You have to demonstrate that he's wrong about LAR.

    LAR shows that when we reach the limits of a technology, the paradigm changes. You say in your post that intelligence can't be recreated by humans because the biology is much more complex than silicon can handle (I'm paraphrasing). But you're missing (or deliberately ignoring) that Ray never said it would be silicon-based. The paradigms will shift. In fact, we're already seeing new technology developing away from silicon.

    You're making the classic mistake of assuming that the future is built on today's technology. You see AI as some Terminator-style steel robot. AI of that era will be every bit as organic as you are. If you find that hard to believe, you're not paying attention to the sciences.

    Ultimately, I believe, based on the science that's unfolding today (and its pace) that Ray's been way too conservative in his estimates. I predict it will be at least 10 years earlier than he predicts, and that's even including new discoveries about the complexity of the brain.

    Here's a tip: Don't rant against the soundbyte or you look like an ignorant fool. Take the time to understand the story BEHIND the soundbyte first. If you do in this case, you will see a whole different picture.

    Gerhard Adam
    Well, someone's been drinking the Kool-Aid.
    Mundus vult decipi
    There's a solid rebuttal.

    We're already in the opening stages of Singularity. People used to be able to zone out at home into World of Warcraft, Second Life, or Facebook, now they can take it all on the road with them due to the increased power of mobile hardware. Look at how you see out in public how people are zoned out and glued to technology, doing most of their communications through, or using that technology to completely replace decisions or calculations they may have made on their own before.

    The dig at "he's just like the homeopathy field" is also pretty obtuse and also pretty judgmental. Yes, he does make alot of money, but I know I have certainly benefited in my life from alternative medicine, particularly supplements, long before their benefits were being recognized by mainstream. Many of these have been dramatically life changing.

    Hank
    You confuse homeopathy with supplements, which may be why you do not understand that the delta between playing WoW and ascension into an artificial intelligence is huge.   You likely buy Ray's books - yayyyyyy capitalism - but it's not judgmental to compare him to the same people who exploit the gullible in medicine for their own cynical gain.   It's fact.

    Another fact; we're not going to be any different by 2045.  I'd bet Ray on that but he won't be alive for me to collect, no matter how young he thinks his 'biological' insides are.
    Never bought any of Ray's books, but I've followed his stories for the longest time over the news sites.

    Homeopathists were pushing some of these "supplements" long before they went mainstream. Now you have mainstream pharmas moving into natural treatments years after they have already been proven to work.

    You two just come across as some grumpy naysayers, resisting change or anyone attempting to better themselves. It's also really arrogant to claim you are somehow so much more intelligent or enlightened than other people by judging them as "gullible."

    The personal dig on Ray (which is common) about him not being around to see Singularity is pretty crass also.

    But thanks, it's this kind of fuel that propels the pioneers to the next level.

    You need to go back to grade school and first figure out the difference between a "fact," and an opinion.

    Hank
    Nooooo, you seem to lack basic understanding of science, technology and medicine.  If homeopathists endorse cute puppies, it does not make homeopathy correct, though many agree puppies are cute.   You seem to think homeopathists peddling snake oil but also selling a natural medicine validates homeopathy.
     
    There is no pioneering work being done by Kurzweil.  He is trapped in the 1950s.  It's all "here we are now"  mumble, mumble about how awesome computers will get "singularity!"
    How the hell do you know what I know or my life experiences are? And you have suddenly judged the body of my work and life experiences next to yours on a scale and somehow think you are sitting on top? That's freakin arrogant. Just like how you dismiss Kurzweil, even though he has a body of work and predictions that show he was accurate. With what? "He's stuck in the 1950s." What does that even mean? The guy created a device to read text to blind people in the 1970s, does that mean he's stuck in the 1970s too? You'd probably be making fun of his idea back then too.

    Your charge against both Kurzweil and homeopaths are hilarious. People on bleeding edge will continue to push the boundaries while you get left behind. They are the ones experimenting and pushing ideas to the forefront while all the naysayers are really the ones stuck in the 1950s.

    But I think you've read too much Quackwatch - quick to dismiss everything but afraid to stand for anything.

    Hank
    Since you believe in homeopathy, you apparently only need to read the first letter of articles and comments to get the full effect.   That must save you a lot of time.   Science certainly has nothing to do with your life 'experiences'.  Data is not relative.

    It's not arrogant to say homoepaths are charlatans - that has been proven for 200 years and not once has any study shown it to work.   Ray is just a misguided old guy who does too much self-experimentation.    Even he can't believe what he says any more, he is just going through the motions for his fans.
    You actually have to HAVE data to form a conclusion. You have zero data on me but obviously have a ton of confirmation bias in yourself.

    For someone who claims that you have such an objective grasp on scientific method, you sure do base most of your arguments on subjective reasoning. "Oh he's just an old guy." "Oh he doesn't even believe himself anymore." (What, you can read minds now?" "Oh he is just a snake salesman."

    Arrogant and crass.

    Gerhard Adam
    We're already in the opening stages of Singularity.
    We're nothing of the sort.  If we're lucky we won't turn ourselves into a third world nation, but this fantasy that humans are "evolving" into some human/machine future organism is simply rubbish.

    In the first place, it is the 21st century equivalent of a technological eugenics program and the assurances from transhumanists simply demonstrates their naivete.  They brush off true complexity as if its simply an engineering problem that will be solved in a couple of decades and demonstrate their utter ignorance of biology in assuming that its a simple mechanistic architecture that they can replicate. 
    Mundus vult decipi
    We don't even know what the neural code is, or if there is one. We don't understand the implications of volume transmission, of why there are 5 classes of dopamine receptors and 12 for serotonin receptors each with variants. We don't know what "information" is with respect to neural transmission. By what warrant can we then make any claim to knowing when artificial intelligence will attain the qualities we seek in it? So I find all such proclamations as to when the Singularity will occur to arise from a strikingly foolish hubris.

    Not sure if someone mentioned this, but in Ray's book "The age of spiritual machines" (published in 1999) 2045 is the date he gives for the singularity. So I think your facts are wrong about that date shifting so much. You also have to understand that predictions on a logarithmic model tend to be more accurate the further out the prediction is on a time line. By my own estimates Ray has been off about 30%, given his timeline he sets forth in the aforementioned book. If his error rate continues to be about that (which is something he has said isn't a problem, for instance he might be off by 10+/- yrs) the singularity is still happening within our lifetimes (well I'm 26 so...).

    Hank,

    I realize this comment is late, but as I'm just learning about singularity theories, I feel compelled to share a thought. That thought is that whether or not Ray is correct regarding IF what he claims will happen, actually does, or WHEN it will happen, you have not offered one shred of scientific evidence to refute that it will not happen at any given date. If you have a book out providing verifiable evidence to refute what Ray says, by all means, point me to it as I would love to see both sides of coin. As it is, I'm still determining whether or not I believe Ray's theory to be plausible (not that it truly matters). Nevertheless, until you can add something of value to an argument which you are so vehemently against (as far as I can tell, you cannot), please refrain from posting complete non-sense.

    Gerhard Adam
    You should know that one can't prove a negative.  The onus is on Ray [or those that profess the same belief] to prove his statements.  Your statement about having a "book out" simply indicates that you're more impressed by arguments from authority than the actual evidence.  Do you think that everything published is somehow "true"?

    I'm still determining whether or not I believe Ray's theory to be plausible (not that it truly matters)
    Welcome to those that want to turn science into a religion.  This is precisely the mindset that it takes. 
    Mundus vult decipi
    Hank
     you have not offered one shred of scientific evidence to refute that it will not happen at any given date.
    This is impossible.  I also can't offer one shred of scientific evidence that Hitler will not emerge from an Antarctic fortress and establish the Fourth Reich in 2045.  I mean, anything at all is possible if the only refuting method is to not prove something may happen in the future.
    If you have a book out providing verifiable evidence to refute what Ray says, by all means, point me to it as I would love to see both sides of coin.
    There's no way that book could be written.  Ray simply says Today + FUTURE MAGIC = Future Singularity.  How can I dispute future magic that has no technology roadmap, no basis in science and a really low level of neuroscience understanding? There's nothing to refute.  People also claim the world will end in 2012. No book can refute it won't happen and it wouldn't sell if they did write it; people buy books claiming goofy stuff might happen but no one will buy one saying it won't because gullible people only buy books saying it will - the rest of the world knows it won't so they don't need to buy a book
    Both points taken. Can you give your opinion on what you would point to as the magic he describes? Do you think it's just the general belief in the exponentiality and extrapolation of where he believes it will lead to, or just a combination of other things he mentions?

    Hank
    No one complains about science/tech optimism - we don't criticize people in 1900 who saw us all taking flying cars to the Moon, for example.  Optimism and fantasy lead to great things that take us out of the old ways. But he tends to attack people who note his belief that because the brain has redundancies they are junk that can just be removed in a digital model is flawed. He certainly lives it and I wrote "Kurzweil has always been an interesting person, a fun guy to sit around and talk with" but his insistence that if people don't subscribe to his vision they don't understand technology or science is exactly wrong; it's because we understand a lot more about technology and the brain than anyone knew in the 1960s that he is so obviously wrong now.
    Herein lies the error in your thinking. You say "Ray simply says Today + FUTURE MAGIC = Future Singularity" which is a misrepresentation of his position.

    In fact, Ray simply says Today's Technology + Law of Accelerating Returns = Future Singularity. This is the point you ignore in everything you have written, both in the article and in the comments. Ray does nothing more than make predictions based on his calculations of those variables.

    What is frustrating is watching someone ignore the actual discussion in favor of the straw man argument. Even if the dates of his predictions were 100 years off, the basic equation is accurate. We are moving at an accelerating pace toward Singularity. Looking more specifically at his data, his case is compelling. Those of us who don't understand your aggressive and emotional arguments wish you would address the basic equation, and not get so stuck on the actual date.

    Hank
     Ray does nothing more than make predictions based on his calculations of those variables.
    and
    Even if the dates of his predictions were 100 years off, the basic equation is accurate. 
    Interesting. So if I write a book predicting the end of the world, you will buy it.  I mean, it doesn't matter if my date is right, eventually the world is going to end, so I remain awesome.   In actual science and technology, if your 'calculation of those variables' proceeds from a bunch of false assumptions, you get fired.  Because he is writing books for gullible people, though, the technology equivalent of astrology readers, I guess it is okay.  It's still silly to not expect any skeptical or scientific community not to call him on it since his predictions are based on a high school level of neuroscience and his high school level of computer technology.
    Gerhard Adam
    ...the basic equation is accurate. We are moving at an accelerating pace toward Singularity...
    There is no basic equation.  The "accelerating pace" is a short term phenomenon that hinges almost exclusively on Moore's Law having no boundaries.  It is complete and utter hype and has absolutely no scientific basis.

    http://www.science20.com/gerhard_adam/transhumanism_ray_kurzweil_and_singularity

    http://www.science20.com/gerhard_adam/blog/computing_speed_limit

    http://www.science20.com/cool-links/microsoft_cofounder_paul_allen_not_buying_singularity-83607
    Mundus vult decipi
    Hank
    And there is a basic physics issue Kurzweil is unaware of (along with neuroscience) - current semiconductors tap out in the next 3 years, electrons will stop moving, so he is counting on some future replacement that will just appear - magic, like I said.
    The Singularity has always been set for the year 2045. Maybe you should go to Mr.Kurzweil's website, KurzweilAI.

    IF a superior intelligence suddenly evolved on the internet, why would it reveal itself? Perhaps it is already there. I don't normally think about communicating with cockroaches and they have been around millions of years.

    Gerhard Adam
    LOL .... Gotta love people that equate themselves with cockroaches.
    Mundus vult decipi
    Glad you enjoyed the joke. Actually, I am the singularity and you are the unter-menschen now.

    I didn't see any mention of quantum computers, or of parallel quantum computation. My dad told me in 1974 or so that we would never have, as they did in the Dick Tracy cartoon, wrist-watch-sized communications devices with full color screens, because there was a lower limit to the size of specific bits of the technology. At age 6, I asked him, "But what if they find something that works better?" He assured me that if it could be found, it would have been already. And then came the silicone computer chip, with transistors too tiny to see, etc..

    As for us not being able to develop intelligent devices more intelligent than ourselves, because we would not be able to tell that they were better, on that point I am quite dubious. If we create a system that can replicate itself, but which tests out minor modifications to specific systems, even differences in component composition, then it would be the same as us reading our own genome, and then deciding, once we have it figured out, which systems we wanted to enhance (such as viral/bacterial immunity), versus ones we might wish to leave alone (skin pigmentation variations). The main sticking point for machines evolving past us is that they have no inherent "desire" to do so. There is no equivalent of machine sex and reproduction. If we create a machine that has as a core component of its programming that at certain points it *will* reproduce itself, that is the equivalent of our biological imperative. If we also give the computer in that machine the desire to 'improve' itself, that is the equivalent of our manipulation of or own genome, epigenome, and physical environment. The only way we would ever have some bizarre future war for superiority would be if the machines had those two things, and then came to the conclusion, as they did in the Terminator series, that the best solution to their own evolution was the elimination of their competition. However, if something becomes superior to us, we would no longer be competition. And that ignores the probability that the 'intelligent' machine would recognise the benefits to itself of cooperation with the 'lesser' species (us included).

    As for us not being able to increase human intelligence artificially, the argument I have seen against this is basically that all we would be able to do is increase our physical memory space, in effect making us have to mentally 'read' articles stored in the artificial parts of our brains. However, if we developed devices that accurately stored information in the same manner in which it was learned (such as was shown in Total Recall), then our 'new' abilities or knowledge would be accessed the same way all our regular memories are. We would 'remember' learning something which we had not physically been present for, rather like how we remember things that 'happened' in a dream. Our mind is set up to run holographic simulations, so it is not far-fetched to believe we will eventually manipulate that to our benefit. I do not care one whit about this Singularity idea, nor its dates. Clearly our sciences are going to be held back by our modern Luddites, and I don't just mean religious fundamentalists (though they have been a thorn for centuries). However, the mass of modernity, the fact of the Internet, the pace of information exchange, all point toward ever increasing complexities of all sciences, regardless of what we might want. We are already modifying ourselves, from implants that have little to do with our brains (artificial hips), to implants that change the way our brain works (deep brain stimulators of area 25 for those with depression). The finer our understandings of biological processes, the easier it will be to alter them artificially. It's just that simple. It sounds too simple, like a religion's dogma, but sometimes Occam's Razor works just fine. There will be no clear line between artificially enhanced people versus not, versus synthetic people. We will advance into becoming something like the Cylons, given enough time and technological advancement. And just as advancements like speech flipped through the entire human species faster than ordinary mating practices would predict, due to its high selection value, so too will synthetic augmentations. Consider how the world is now linked with cellular telephony and satellite television/Internet *everywhere*, even in the nations where they still don't have clean drinking water, or enough to eat. The desire to share information is greater than the desire to eat, drink, or screw, because it is so powerful. And there is no reason why we can't combine eating, drinking, screwing, and sharing information.

    As to the power consumptive needs of newer tech, a look at televisions reveals a *downward* turn in their power use, due to a change in the technology of their screens and other basic components. There is reason to believe that, just as battery efficiencies continue to advance, and as solar generation continues to increase in efficiency, those things plus other technical advances will lower the amount of power needed to do the same amount of 'computation.' We are finally understanding that what we call 'human intelligence' is really just elaborations and adaptations of things animals around us already have. If we manipulate any of the variables involving our elaborations, then we will affect 'intelligence,' whatever that ultimately is. If we develop a means to send thoughts via technology as pseudotelepathy based on implants able to read our thoughts and send and receive data, it is not inconceivable that there would come to be some level of different sorts of hive-minds, or even a sense of always being linked in, just not having our 'chats' turned on. Our largest impediments to all of these changes are some of the things that make us human, our love of eating, drinking, and having sex (and using recreational substances). Until we can subjugate/satisfy our physical needs entirely, and make this something *all* people can easily do, those needs in the majority will continue to determine the directions and speed in which our technology develops.

    Hank
    Philosophically, you are right - but only because you can't be wrong.  Obviously anyone can say 'my dad said we would never have a computer in a phone and now we do' as unbridled faith.  Kurzweil certainly does.  But that is a religious roadmap and not a science or technology one just like if I claim I will go to heaven.

    Look, some day I think we will no longer need an organ donation card or a transplant waiting list. I believe we will be able to grow new body parts.  Does that make me a visionary just by claiming it?  In reality, we have no technology that could create enough stem cells, or a complex scaffold, to do any of the important organs yet, even though we can do small, simple ones - we don't even yet have the technology to know how to make the technology to make the roadmap.  So my creating a date would be just silly, it is a guess to sell some books and tickets to conference, like any good prophet does.
     as battery efficiencies continue to advance, and as solar generation continues to increase in efficiency
    Solar has been around for 50 years.  And for 50 years we have been told it will power the future. Tell me what meaningful advance has occurred in solar power efficiency.  None at all - the only meaningful advance has been that Chinese cheap labor and no environmental policies have made them able to sell panels cheap and governments are willing to spend billions subsidizing them - that is no advance.  And batteries are not more efficient either.

    Hope and 'computers got better for 50 years so they will always get better' is not a plan.  The gap between the vision that Kurzweil laid out and his knowledge of neuroscience is a grand canyon - it really is a black box with magic inside.   If quantum teleportation becomes commonplace in the future, we are not going to call Harry Blackstone the father of it.
    Gerhard Adam
    As for us not being able to develop intelligent devices more intelligent than ourselves, because we would not be able to tell that they were better, on that point I am quite dubious.
    You can be as dubious as you want, but you can't confirm it. 
    The main sticking point for machines evolving past us is that they have no inherent "desire" to do so.
    If that is true, then they're already more intelligent that us.  I won't go any farther, because that concept is about as stupid as humanity can get.
    Mundus vult decipi
    I think the author of this article misses an important point. He says that Moore's Law + "magic science" is the basis for the AI prediction. But that is not the case. Instead the idea is very simple: you don't have to solve something algorithmically if you have the power to solve it by brute force.

    For example, if you had the ability to scan a brain fully then you could copy it even if you didn't understand it. Or alternatively, you could run billions of possible AI algorithms and sort through which ones are most successful without having to understand the merits.

    Of course the most classic example is the game of chess. It is not yet fully solved, but Moore's law will allow us to fully solve it (meaning all possible games are played so that a computer can be absolutely sure about what the best move is at every point). Instead of building up a bunch of algorithms (e.g. "develop your knights early", "control the center", "maintain a good pawn structure") a computer won't need any algorithm at all to win all the time at chess -- it will simply know all the possible games via brute force.

    There are lots of science problems, mathematical problems, etc. where the likelihood of a closed form algorithmic solution is low, uncertain, or even proved impossible; yet a brute force solution will work.

    Anyway, the point of Kurzweil bounding the point at which super-AI is possible is based on the point when brute force alone can solve the problem of AI based on when we would be able to scan a brain fully and copy it (maybe with an equivalent software or electronic circuit) without any need to understand it.

    Kurzweil does have an optimism problem though. It is not in his date, but in the fact that he thinks that just because something is possible that it will become common. For example, it is certainly possible TODAY to have a flying car, it is possible TODAY to implant a lot of electronics into our bodies, it is possible TODAY for humanity to provide food to everyone, it is possible TODAY to take hormones that turn you into a superman, but we simply haven't chosen to make any of these things common or societally acceptable.

    Gerhard Adam
    Anyway, the point of Kurzweil bounding the point at which super-AI is possible is based on the point when brute force alone can solve the problem of AI based on when we would be able to scan a brain fully and copy it (maybe with an equivalent software or electronic circuit) without any need to understand it.
    That's the "magic" part.  The fantasy that the brain can be simulated by "brute force" is laughable.  Coupled with the notion that it can be "copied" or replicated without the need to understand it, is simply foolishness.  That's why the entire proposition is magical thinking.
    Mundus vult decipi
    Why is it laughable and magical?

    Its all well and good to say Kurzweil's predictions are "silly" but not once in the article did I see a logical reasoned explanation (or any explanation at all for that matter) as to why the author believes Kurzweil's predictions are silly. I am all ears, please tell us why Kurzweil is wrong.

    Hank
    Because everyone understands the brain is complicated but him.
    So you think the brain is too complicated to be simulated. What evidence do you have for this?

    Gerhard Adam
    LOL ....
    Mundus vult decipi
    ???

    Gerhard Adam
    Simple ... Kurzweil has no science to back up his claims.  They are based on the erroneous assumption that somehow progress [as he defines it] will continue at the same rate and in the same vein so that he can reach his "target date".

    There isn't a single shred of evidence to believe that his claims have the slightest chance of coming true.  Therefore, it isn't on us to demonstrate where he's wrong.  It's on him to demonstrate why he's right, and he simply doesn't have the data.  What he does publish is pure speculation with no scientific backing.
    Mundus vult decipi
    Kurzweil has uncovered a remarkably persistent and stable trend in the exponential growth in technology. Yes it is true that past trends do not guarrantee future trends. But using a Bayesian
    probabilistic approach, such a trend would be expected to continue into the future with high probability. If you have read his book the singularity is near he has attempted to demonstrate and give evidence for his views. What specifically about his evidence (that he has laid forth in detail in his writings) do you find unconvincing?

    Gerhard Adam
    What specifically about his evidence (that he has laid forth in detail in his writings) do you find unconvincing?
    All of it.  The simple fact that he has been most wrong in his predictions, beyond those that are relatively simple trend analysis.  Kurzweil has demonstrated that he loves to speculate without any regard for the actual capabilities and this is a common problem when the only method of prediction is simplistic extrapolations.
    Mundus vult decipi
    I repeat, what specifically about his evidence (that he has laid forth in detail in his writings) do you find unconvincing? Which of his arguments do you find particularly unappealling? Have you even read the book?

    Gerhard Adam
    Perhaps the simplest and most obvious is the reliance on Moore's Law.  There is no such thing and the dependence on it holding true is a problem.
    Mundus vult decipi
    Excuse me? Seriously, did you even glance at the book "The Singularity Is Near"? If you had you would have noticed that Kurzweil specifically acknowledges that Moores law will come to an end and his theories in no way depend on it holding true. Stop embarrassing yourself. Read the book. When you're done reading it come back and give your opinion of Kurzweils ideas. Unbelievable.

    Gerhard Adam
    OK, I get it.  You're a true believer.  I read the book years ago, so I don't particularly care what the specific argument is that you're looking for.  There are plenty of more recent articles and discussions and not one of them as a modicum of reality associated with them.

    The fact is that Kurzweil simply argues that there will be a sixth paradigm that occurs after Moore's Law to replace the effect.  You can easily see the projections here, the majority of which are utter nonsense.
    http://www.kurzweilai.net/the-law-of-accelerating-returns

    If you want to argue about a specific point, then feel free to bring it up.  However, there isn't going to be a singularity.
    Mundus vult decipi
    This article is putting forth outright lies as the truth, and I find it outrageous. Kurzweil has always been very firm on the Singularity happening in 2045, and you are taking his other statements out of context.
    For example, at one point he said "within 25 years computer intelligence will be greater than that of human intelligence" That was not a prediction about the singularity, rather it was related to another prediction he made about the exponential increase in brain-scanning quality and how much data we get out of them.

    Gerhard Adam
    What difference?  Wrong is still wrong.
    Mundus vult decipi
    I would like to point out that this analogy itself is incredibly flawed. A CD is a static copy of a set of data, whereas the brain is a constantly evolving dynamic set of ever-changing points of data, not to mention all of the complex data on the chemical carriers and the like that is also completely unique to every person and also constantly changing. Therefore you have no way of knowing how shortcuts would affect it. The whole premise of your argument is therefore fundamentally fallacious.