Bookmark and Share

Alex Bäcker's Wiki / The Age of Spiritual Machines, by Ray Kurzweil
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

The Age of Spiritual Machines, by Ray Kurzweil

Page history last edited by Bruce A. McHenry 13 years, 3 months ago Saved with comment

It is perhaps fitting that on my way to and from Future Camp (aka Singularity Point) at Burning Man I read this book. The book has several interesting passages, as the other book by the same author I have reviewed, and makes for entertaining reading. Here, I will limit my comments to the chapter on Kurzweil's 1999 predictions for 2009, as those can be verified against what really happened. It turns out that he was overly optimistic. Despite writing himself that people tend to overestimate progress in the short term, he fell victim to the same trap. He got a few things right ('most routine business transactions will be conducted between a human and a virtual personality', and 'computer displays have all the display qualities of paper', for example), but here are 25 of the many predictions that Kurzweil got wrong:

 

  1. People typically have at least a dozen computers on and around their bodies. I am in the high-tech industry in the most technologically advanced state of what is perhaps the most developed country on Earth and still I only have my iPhone and my laptop. I have never met anyone with a dozen computers on and around his/her body.
  2. PCs have no moving parts --solid state drives are here, but are still far from routine.
  3. Most portable computers do not have keyboards. There were more than 500M computers sold in Q1 2009, of which more than 30M were laptops, vs. just about 16M smartphones. So the great majority of portable computers still have keyboards. And of those that don't have traditional keyboards, like iPhones, the great majority has a digital touchscreen keyboard.
  4. Human musicians routinely jam with electronic musicians. Not last time I heard. Of course samplers have been around for years, but I have yet to see automated compositions make the charts.
  5. The majority of text is created using continuous speech recognition --not any interesting text that people read, anyway.
  6. Computer displays built into eyeglasses that project onto the user's retinas are also used  --none that I have seen; Wikipedia states that 'no VRD-based system has yet reached operational use and current military HMD development now appears focused on other technologies'.
  7. Computers routinely reliably identify their owners from their faces --none recognized me yet, but then again perhaps I'm a chameleon.
  8. 3-D chips commonly used --a Google search for 3-D chips yields 3-D chip design challenges as its top result, and a question about whether they'll be able to take Moore Law's beyond 2020 as the second result.
  9. $1k PCs can perform a trillion calculations per second.  My Mac can only perform 0.5% of that,  and cost more than 1,000 1999 dollars.
  10. Students interact with their computers primarily by voice and by pointing with a device that looks like a pencil. Not the students I know.
  11. Preschool and elementary school children routinely read using print to speech technology. I don't know one who does.
  12. Translating telephone technology is commonly used. I'm waiting for this one! I know it's coming, but even text to text translation is still quite poor.
  13. Virtually all communication is encrypted --encrypted email is rare and its UI arcane.
  14. Haptic technologies allow people to touch and feel objects and other people at a distance. Call me insensitive, but I haven't felt any yet.
  15. People have sexual experiences at a distance with other persons as well as virtual partners. Call me old-fashioned...
  16. Phones routinely include high-resolution, real-time moving images of the person on the other end. iPhone 4 came out in 2010, a year behind Kurzweil's schedule, but is still far from commonplace despite decades of predictions to the contrary.
  17. The ten years leading up to 2009 have seen continuous economic expansion and prosperity...oops, I hope Kurzweil did not put his money where his mouth was. Stock market indexes such as NASDAQ and the Dow Jones Industrial Average were well below levels of 1999 throughout 2009.
  18. Intelligent roads are in use. I guess they took the road less traveled.
  19. A company west of the Mississippi has surpassed a trillion dollars in market cap. The most valuable public company on Earth is Exxon. Its market cap less than half a trillion dollars (less than 50% of Kurzweil's prediction), and it's headquartered East of the Mississippi, in NY.
  20. There is a growing neo-Luddite movement. Google Trends reports that neo-luddism does not even have enough searches to merit entering its index, and neo-Luddite does not fare much better.
  21. The underclass is politically neutralized through...the generally high level of affluence. Last I read, unemployment were at decades-high record levels.
  22. The creation of music has become available to persons who are not musicians. I have not heard of any...has any of these reached the charts?
  23. Cybernetic authors are emerging. None that I read or topped the NY Times Best Seller list--a search for cybernetic authors still yields mostly Kurzweil's own book.
  24. Bioengineered weapons constitute the greatest threat to our national security. Of course one can speculate about what the greatest threat is, but there has been no bioengineered attack on the US to my knowledge. Predictions such as these are dangerous, as they can cause our governments to spend billions on preventing something that has never happened and has little chance of happening soon, whilst millions die preventable deaths that nobody is spending on.
  25. Telemedicine is widely used. Most doctors don't even use email!

 

It's not enough to point out a few research studies in a few isolated labs around the world --research in all of these topics was ongoing at the time of Kurzweil's writing, and there is a long road from demo to mass market. Kurzweil made specific predictions about how commonplace these technologies would be.

 

In sum, Kurzweil got a lot of it wrong. Not just plain wrong. He got multiple predictions off by more than three orders of magnitude. And this with his closest range predictions --only 10 years after his writing. Why? Perhaps because while a few quantitative measures do grow exponentially, a lot of the things that affect life do not. We travel in planes pretty much the way and at the speed we did decades ago. Same with our cars' speed. Most of our movie theaters are similar to what they were. The food most of the world eats is still similar to what it was decades ago. And education is still effected largely manually by teachers in classrooms. Our bathrooms have not changed much, either. Our clothing changes styles, but is still made of inanimated non-electronic materials. My grandmother can still understand what I do for a living. With the exception of a few arenas, such as computing and telecommunications, life really has not changed all that much. Change is limited by man's willingness and ability to change. People's taste gets fixed during their youth, and so dramatic change must await a new generation. And generations turn over linearly over decades, not exponentially. And whilst individuals can make changes fast, large organizations, such as multinational corporations and governments, have greater inertia and change much more slowly. And a lot of our life is guided by governments and corporations. Even in arenas such as computing where exponential trends have emerged, performance on most complex tasks is subject to bottlenecks caused by the worst-performing of multiple elements. For example, not only is the computer I bought this year not a whole lot faster in CPU speed terms than the one I bought in 1995, but it's performance these days is not limited by CPU speed, but rather by bad software, bus speed, memory or hard disk speed.

 

In the time since his mistaken predictions, Kurzweil and his followers have created conferences and even a university focused and predicated on those predictions. Having attended a couple of them and spoken at one of them, I know these failures are not often discussed. If they are to help to bring the world into a better future, I believe they must start by examining the failed predictions, and the reasons behind them.

 

As exciting as Kurzweil's predictions are to fathom, I'm afraid we'll have to wait a little longer to make his science fiction fact.

 

P.S. Kurzweil has responded to the criticism, and by his count the overwhelming majority of his predictions were correct. In fact, he cites a single one as wrong, and says that one was tongue in cheek (he did not specify which). However, I believe the average reader would score them very differently. Here is a specific example used by Kurzweil to illustrate his point: "The correct prediction was “Personal computers are available in a wide range of sizes and shapes, and are commonly embedded in clothing and jewelry…” When I wrote this prediction, portable computers were large heavy devices carried under your arm. Today they are indeed embedded in shirt pockets, jacket pockets, and hung from belt loops. Colorful iPod nano models are worn on blouses as jewelry pins or on a sleeve while running, health monitors are woven into undergarments, there are now computers in hearing aids, and there are many other examples. The prediction does not say that all computers would be small devices, just that this would be "common," which indeed is the case.. And “computers” should not be restricted to the current category we happen to call “personal computers.” All of these devices – iPods, smart phones, etc. are in fact sophisticated “computers.” By a reasonable interpretation of the prediction and the current reality, it is correct, not “false.”, wrote Kurzweil. The corresponding original prediction was:"Personal computers are available in a wide range of sizes and shapes, and are commonly embedded in clothing and jewelry such as wristwatches, rings, earrings, and other body ornaments." Now ask the average reader if they'd say that was true. I'm an inventor in California in the tech industry and don't have a single computerized wristwatch, ring, earring, clothing item or body ornament. Nor do my friends. So to call that prediction true is a far stretch from reality, and invalidates any count by Kurzweil himself.

 

P.S.2. A comment by Tim McCune, whom I have a lot of respect for:

I enjoyed http://alexbacker.pbworks.com/The-Age-of-Spiritual-Machines%2C-by-Ray-Kurzweil and wanted to respond to some of your points.

"Human musicians routinely jam with electronic musicians. Not last time I heard. Of course samplers have been around for years, but I have yet to see automated compositions make the charts."

Actually, a lot of composition software like Cake Walk goes a long way toward that.  Also wonderfully relevant: http://nlp.stanford.edu/courses/cs224n/2009/fp/5.pdf

"$1k PCs can perform a trillion calculations per second.  My Mac can only perform 0.5% of that,  and cost more than 1,000 1999 dollars."

>From http://en.wikipedia.org/wiki/FLOPS discussing a card released almost a year ago:
"AMD Radeon HD 5970 Hemlock, the first high-performance graphics card with dual GPU from AMD, which breaks the 1 TFLOPS mark in Double precision floating-point. It can reach speeds of 4.64TFLOPS (sp) and 1.09TFLOPS (dp) respectively when running at 2 * 725 MHz.[38] Price per GFLOPS is slightly inaccurate as it includes only the cost of the card ($640), which will drop in time."
Also, CPU clock speed != operations per second.  I have a hunch that your Mac can do much better than you're giving it credit for.
http://en.wikipedia.org/wiki/Million_instructions_per_second

"Virtually all communication is encrypted --encrypted email is rare and its UI arcane."
Sigh.  I'm painfully aware of your issues with encrypting your e-mail. :)  However, assuming we're talking about non-verbal communication, cell phones actually account for a good deal more communication than e-mail if you believe Randall Munroe's cartographic skills.
http://imgs.xkcd.com/comics/online_communities_2_large.png
And your phone's communication is encrypted
http://en.wikipedia.org/wiki/Gsm
even if it's not done very well.
http://www.engadget.com/2010/01/15/3g-gsm-encryption-cracked-in-less-than-two-hours/

"Telemedicine is widely used. Most doctors don't even use email!"
True, but 39% of physicians now email, secure message, or instant message their patients, so it won't be true for much longer.
http://americantelemed.blogspot.com/2010/03/adoption-of-telemedicine-accelerating.html
My grandmother had quadruple-bypass surgery a few years ago, and her surgeon checked in on her daily by using a robot that he could wheel around the hospital, from the comfort of his home computer.

 

P.S. 3. Feb 2011 update: 

Kurzweil's reply: http://www.kurzweilai.net/predictions/download.php

Much of this is subject to interpretation. For example, Kurzweil cites the prediction that 'Preschool and elementary school children routinely read using print to speech technology.' as correct because he works with kids with disabilities that do. As I wrote, I don't know one who does. So I guess you'd have to ask him what he meant by routinely. My interpretation is that meant that this is common, as in a majority, or a number close to a majority, of kids does this. Not as in 'most people have never heard of it'. But I guess you could interpret it as '2 or more children do it as part of their routine'.

 

Kurzweil counts his prediction of electronic musicians jamming as correct because "Such ―auto accompaniment‖ software is 

also built into home digital keyboards". This was built into my inexpensive musical keyboard when I was growing up in the 1980s, so if his prediction made in the 90s counts that as a prediction, then it wasn't much of a prediction.

 

Some of the predictions that Kurzweil admits were wrong are those with clear cut quantitative predictions, such as the one below:

'PREDICTION: Keyboards still exist, but most textual language is created by 

speaking.

ACCURACY: Wrong

DISCUSSION: Text creation by speaking is certainly common: Dragon Dictate from 

Nuance is used by millions of consumers, and Dragon Dictation for the iPhone, iPod 

touch, and iPad is one of the most popular iPhone apps. Nuance Technologies, Dragon 

Dictate‘s parent company, now has more than $1 billion in revenue and has consistently 

experienced 50% annual growth. However, it is not the case that ―most textual language‖ 

is created this way. This will probably not be the case in several years either, but may 

plausibly become the case within ten years.

 

In contrast, predictions that concern advances much farther away than speech to text yet were worded in a more ambiguous or modest way are marked as correct:

"PREDICTION: Research has been initiated on reverse-engineering the human 

brain through both destructive scans of the brains of recently 

deceased persons as well as noninvasive scans using highresolution magnetic resonance imaging (MRI) of living persons 

and animals.

ACCURACY: Correct"

 

Thus, the precise wording of the prediction may be more important than its substance in determining its correctness.

 

  • David Vorriccelli 
    ‎@Alex Backer, the list in link you've posted just plain wrong on a lot of the topics:

    #4. Pic a top 40 hit. It's electronic. All of the beats, and most of the vocals are digitized. 

    #10. My friend's 1year old has a mass produced book with astylus that reads aloud wherever it touches on the page. Granted, he's not the one using it... it's his dad who gets the kick out of it.

    #13. HTTPS?

    #15. Chat-roulette = sex at a distance

    #16. Teleconferences? My boss uses them every day.

    #25. Davinci machines? 

    (unfortunately, i'm still getting used to the facebook shift+enter feature... please excuse the multiple updates. )
    6 hours ago · Like

 

 

  • Alex Bäcker 
    David, first, yes, facebook's change of the ENTER meaning is ridiculous.

    #4. Being digitized does not mean that it's an automated musician that's composing the songs or jamming with human musicians. Yet. Remember I don't disagree w/the direction it's heading, just w/the speed.

    #10. As you said, kids are not yet reading this way. It's not that they couldn't.

    #13. That's not the majority of electronic communications.

    #15. Well, if you will count that, then it was present before the predictions via phone sex.

    #16. Teleconferences don't involve moving images. Videoconferences do. They exist, and I use them, but phone calls still don't routinely include video. This is about human behavior, standards and bandwidth, not about technological capability. As I mentioned in the post, which talks about a year that's 2 yrs ago, this one is not far behind, though.

    #25. I presume you are talking abou this?http://www.intuitivesurgical.com/products/ I doubt Kurzweil was talking about medicine from centimeters away when he spoke about telemedicine, but yes clearly long-distance surgery is next.

    Overall, K's predictions are pretty good, but there is a clear bias in the errors: he tends to be overly optimistic. Which gets him press.
    about a minute ago · Like

 

 

Comments (2)

Tim said

at 10:17 pm on Dec 31, 2010

"Human musicians routinely jam with electronic musicians"

Check out http://www.ujam.com/

Alex Backer, Ph.D. said

at 5:23 pm on Feb 18, 2011

Tim, re my Mac's speed, that link you gave on Wikipedia for MIPS gives this as the fastest computer in 2009: Intel Core i7 Extreme 965EE 76,383 MIPS at 3.2 GHz 23.9 2008
That's 7.6% of the trillion IPS predicted even if my Mac had that --more than I had estimated, but still a lot less than predicted by Ray.

You don't have permission to comment on this page.