Kurzweil Reveals an In-Depth Analysis of His Predictions for 2009 in Letter to IEEE Spectrum

In a recent letter written to John Rennie responding to his recent critique of Ray’s predictions, Kurzweil defended himself and his predictions, and most importantly, linked to this. This huge document is over 150 pages long and packed with cool images and facts.

Kurzweil hits back at Rennie:

While I appreciate some of the things John Rennie has to say, his review of my predictions is filled with inaccuracies, including misquotes of mine, and misunderstandings of the meaning of my words and the reality of today’s technology. For starters, he takes note of my point about selection bias, but his entire article suffers from this bias. While he acknowledges that I wrote over 100 predictions for 2009, in a book I wrote in the late 1990s, he only talks about a handful of them. And he persistently gets these wrong. He writes that I predicted “widespread, foolproof, real-time speech translation.” We do in fact have real-time speech translation in the form of popular phone apps. But who ever said anything about “foolproof?” Rennie just made that up like a lot of the factoids in this article. Not even human translators are foolproof. Apparently that has now been removed from the online version.

It’s true that the only way to really figure anything out is look at each prediction one by one, as Ray has now done. I haven’t read the analysis yet but it looks very impressive. Here’s the punchline on page 5:

As I discuss in detail below, I made 147 predictions for 2009 in ASM, which I wrote in the 1990s. Of these, 115 (78 percent) are entirely correct as of the end of 2009, and another 12 (8 percent) are ―essentially correct (see below) — a total of 127 predictions (86 percent) are correct or essentially correct. Another 17 (12 percent) are partially correct, and 3 (2 percent) are wrong.

Nice detail.


  1. “It’s true that the only way to really figure anything out is look at each prediction one by one, as Ray has now done.”

    One could do a random sampling.

  2. Jay

    Most human beings are just too stubborn to learn to appreciate people in their own time.

    Rachel Maddow just did a ‘Obama did 85% of things on his agenda’ thingy:


    But apparently people are too dumb to give Obama credit for everything that he *did* do.

    Instead, they slam him for not fixing the economy yet.

    It’s the same with Rembrandt, Semmelweis and Kurzweil.

    Nobody will ever be appreciated in his own time. Period.

    I’m glad I’m part of the fraction of the percent of all people that was willing to pull their heads out of their asses before the ripe old age of 30, after which the stiff minded are unable to change their thinking patterns.

    Oh, and I’d also like to wish Michael and all the readers of this blog a merry Christmas!

  3. Panda

    Thank you to Mr. Kurzweil! Happy holidays!

  4. disruptech

    Anyone who creates disruptive technologies will be appreciated in his own time.

  5. my cpu

    One of Kurzweil’s most important and true statements:


    Once a technology becomes an
    information technology, it becomes
    subject to the law of accelerating returns.


    When you turn material things into bits, or just their control and manipulation, it speeds up development tremendously and makes advances possible that were inconceivable and impossible before. We’ve all seen this happening during our lifetimes and we’re still just getting started.

    There should be a word for this “when a technology becomes and information technology”. Perhaps “when it gets computationalized or infotechnologized”?

  6. Jim

    This looks like more of an exercise in semantics than a real assessment of his predictions.

    Take for example: “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.”

    His argument is essentially ‘iPods and smartphones fit in your pockets’, but that’s not even remotely what his prediction was. iPods, smartphones, and computers are not commonly embedded in clothing, wristwatches, rings, earrings, or other body ornaments.

    What he was really saying was “computers will be so small and cheap that they’ll be in everything”, but they aren’t. The second half of this prediction is totally false. And “there’s an expensive designer shoe with a microchip!” or “some research company says there are 500 million wearable microchips!” is not a validation. None of this is common, or even close. If you take a random person off the street and search them for microchips embedded in any clothing, you won’t expect to find anything. So Kurzweil just talks about smartphones for a while until we forget that he’s really just stalling for time until this prediction turns correct.

    Let’s also look at “People typically have at least a dozen computers on and around their bodies, which are networked using ‘body LANs’ (local area networks).”

    Essentially correct? Really? This is one of the most blatantly untrue things you could ever say. First, Kurzweil lists just six devices (one of which is just a bluetooth headset – techically a computer, but not exactly something I do any ‘computing’ with), plus “various communication devices”. I’m not exactly sure what those are supposed to be since he already listed a cell phone. Maybe my modem?

    So we have six or seven real things that SOME people MIGHT have all at once but usually don’t, and even then they aren’t usually all LANed together (but that’s not important, apparently). Kurzweil then realizes that six or seven is less than twelve, and for some reason wants to explain that his number is a bit low, so he reminds us that processors have more than one core sometimes, which counts as more computers! Wow! The future is amazing!

    He goes on to say “One could argue that it is typical already, but it will become very common within a couple of years.”

    What the hell is this? ‘One’ could argue? The whole point of this document is for Ray Kurzweil to argue, not some ambiguous pronoun. He’s effectively admitting that he’s wrong here, and then begging for a few more years to be right.

    Another Kurzweil maneuver: “most” apparently now means “greater than zero percent” or “any”!

    Seriously, what a joke. He minimizes his predictions, exaggerates the present state of things, and somehow we’re supposed to believe that he’s been right all along.

    And yeah, accelerating returns. That’s happening. OK. I get it. It’s amazing. I understand it, and it still blows my mind. Computers are rad and the future might be awesome. But these specific predictions are still terrible.

    • Arie

      It’s pleasing that he did spend considerable effort to review his predictions with lots of examples and references, but his arguments are more or less what i expected.

      Consider this one: “most users have servers where they keep VR environments”.

      Then he cites the Nintendo Wii as an example?
      That means I already had a “server” with a “VR environment” in my possession 15 years ago, before he started writing TAOSM. Seriously, people were shooting 3D ducks on their Sony Playstation console with a lightgun at the time.

      He even consideres the prediction of deployment of “flying weapons the sizes of birds” as essentially correct, because of some program for miniature drone research. So even if a technology arrives on the scene a decade late, he still thinks he’s essentially correct because it might have been based on some very crude prototype in a lab somewhere in 2009.
      I hope for Kurzweil that his prediction of cure-all bloodstream nanobots is not “essentially correct”, because if it is he might miss his “bridge” to indefinite lifespan and die.

    • Jim, if you liked the first part, you should definitely read the whole thing.

      • Jay

        Why, what’s in later parts that so great about the document?

        I randomly scrolled through a few predictions in the whole document.

        @Jim: Why would more cores not count as multiple computers?

        I think it’s quite clear that the future is all about multicore computing. We won’t see many benefits from the shitty 4 or 6 cores we got now… but 16 cores are already on the horizon.

        Not that it matters a lot, because gfx cards blow multicore CPUs out of the water by a factor of a few dozen. And in the end it’s really all about having raw computational power.

  7. acceleratingletdowns

    Some of these essentially or partly correct ones will still be essentially or partly correct in 2020. Come back to this blog in ten, and I can say “I informed you thusly”.

  8. Jim

    @Michael: I did write my post based on just the first few entries (mostly because I was so incensed by what was going on that I needed to say something), but I did continue to make sure that my beef with him was applicable to the whole document. Really, it’s not about those particular arguments but just the way he argued.

    @Jay: If I say, “Hey, let’s have a LAN party at my house, because I have six computers!” and I’m referring to the fact that I have a six-core processor, everyone is going to be really pissed when they show up and there’s only one thing that they consider to be a computer.

    • Jay

      I think you’re trying a little too hard to let Kurzweil’s predictions be false.

      Just because you don’t agree with his side of the extreme, doesn’t mean you have to position yourself at the other side of the extreme.

      Nothing personal. Happy new year.

      • Jim

        I’m more desperate than anyone I know for Kurzweil to be right. If there’s bias going on, it’s definitely in his favor. But I don’t feel that simply declaring all these predictions to be correct will make a difference in my life, or speed up the arrival of the Singularity, or do anything remotely productive.

        I propose the following: let’s individually assess each of Kurzweil’s predictions and compare scores. If our scores are similar you can at least agree that I’m not as biased as I claim, and if they’re dissimilar we can use it as a starting point for debate on each prediction.

  9. Dave

    Why is everyone paying so much attention to what is or is not popular? Isn’t it just about what is possible?

Leave a Comment

Awesome! You've decided to leave a comment. Please keep in mind that comments are moderated.

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>