How my predictions are faring — an update by Ray Kurzweil
October 1, 2010 by Ray Kurzweil
How My Predictions Are Faring | Overview
In this essay I review the accuracy of my predictions going back a quarter of a century. Included herein is a discussion of my predictions from The Age of Intelligent Machines (which I wrote in the 1980s), all 147 predictions for 2009 in The Age of Spiritual Machines (which I wrote in the 1990s), plus others.
Perhaps my most important predictions are implicit in my exponential graphs. These trajectories have indeed continued on course and I discuss these updated graphs below.
My core thesis, which I call the law of accelerating returns, is that fundamental measures of information technology follow predictable and exponential trajectories, belying the conventional wisdom that you can’t predict the future.
There are still many things — which project, company or technical standard will prevail in the marketplace, or when peace will come to the Middle East — that remain unpredictable, but the underlying price/performance and capacity of information is nonetheless remarkably predictable. Surprisingly, these trends are unperturbed by conditions such as war or peace and prosperity or recession.
— Ray Kurzweil
Please download Ray Kurzweil’s full paper “How My Predictions Are Faring” (pdf) here.
Comments (15)
by 'eldras
Re: War vs. Peace
Robert Jungk’s masterpiece “Brighter than 1000 suns” recounting Loas Alamos construction of the atomic bomb states no fundamentally new physics was dont there, despite 000′s of world class scientsists being billeted together.
by DevilDocNowCiv
Ray,
One correction-the correctly predicted trajectories of tech improvement haven’t -in an important way-taken place in times of war and peace.In the US, wars that we have engaged in were wars of choice, that didn’t interrupt R&D-they even spurred it, via increased DARPA funding. So-no wars in core Europe, China, Japan, Singapore, and Australia-and only for brief periods in Israel. That tiny country merits inclusion, because of the contributions of many of its citizens. And the US just had wars of choice that didn’t hurt R&D.
It isn’t that the parts of the world that suffered wars or other causes of national development pause don’t morally matter. Rather, that the tech improvement didn’t have to somehow mysteriously persist during a time of bombed cities and mass hysteria. Only now, in the 21st century, the only expressions of cross border conquest are within the borders of recently freed captive nations of a now defunct empire. The descendant root state of that empire has invaded and exerted conquest and non-conquest hegemony over several of these prior captive states, but unlike Hitler and Czechoslovakia, this doesn’t look like its going to spread outside of the prior captive nations. Even some of them have been admitted to NATO, and others are coordinating with us and the Eurozone. So, most of the developed world will probably continue able to engage in tech development, and the predicted course of that development will, hopefully, accelerate as the ideas available on this site and others are developed-exploring the use of quarks as components of logic gates, for example. And if social dynamics or war do interrupt development in the US and Europe-the Chinese will race the Japanese, and we’ll see if the AI period preceding Singularity turns into a beatific time of plenty, or a different version of augmented reality.
by Mike Block CPA
I cannot begin to thank Ray Kurzweil for his inventions, books and papers. If does not remotely matter whether an individual prediction was accurate, much less accurate by date.
Before Kurzweil, I never saw anyone even suggest that the growth of man-tool-computer power had been increasing at an accelerating rate, for more than 5,000 years. Before Kurzweil, despite my 50+ years of heavt computer experience, I certainly never heard that we were all heading towards having 1,000 computers, at a cost of around $1,000. This suggest that the future acceleration of our combined brainpower will be going nearly straight up, especially as we will live far healthier lives for far longer.
Now we have this incredible predictions paper, which would be far better if it only had that combined brainpower chart. Thank you again Ray, for documenting why the best is yet to come and it is coming so fast.
by Cazbot
Some of your predictions are a couple of years off, like the eye glasses and virtual reality pervasiveness. Do you think this slight margin of error 2 or 3 years or so will be compounded over time? I know you rarely say specifics for certain, is that because its hard to gauge the willingness of the market to capitalize and take risks on new tech, or resolution and predictability of your data sets?
by Jos Smit
I just read the PDF. It’s interesting reading, and I can only confirm that predictions have largely come true which is a well done achievement.
But most interesting, I think, are the graphs at the end. Although I’m somewhat disappointed in the limited amount of graphs, I think there are many more graphs to support the law of accelerating returns.
Anyway, many of the graphs follow more or less a logarithmic growth, some very accurately. However I think it’s also interesting to look at the trends on top of the logarithmic growth.
The graphs “Magnetic data storage” and “Internet data traffic” and “Internet backbone bandwidth” and “total bits shipped” seem to follow a logistic-like curve. Appearently there is an inhibiting factor that increases over time. For the magnetic storage it’s probably a physical limitation, and for the others it may very well be an economic factor.
There are also graphs that seem to accellerate faster than exponential growth; “DNA sequencing cost”, “manufacturing labour productivity”, “US patent applications”, “photovoltaic production”, “calculations/second/$100″..
I think it’s very interesting to look at all these trends separately and try to explain them. And even where we can’t explain them I think for extrapolating the graphs, taking these trends in account will improve the accuracy of all long term predictions.
Happy to donate my two cents,
Jos Smit
by josdorpjossie
Sorry for my poor english. But I hope my intention will be clear.
by sblack
Hello, have you checked out Google as a simple search yet? http://www.google.com/search?sourceid=chrome&ie=UTF-8&q=personal+flying+machines Let us know what level of detail you’re looking for… ;) -Sarah, Managing Editor
by David
KurzweilAI Editors, Writers: I’m interested in this particular prediction:
Prototype personal flying vehicles using microflaps exist. They are also primarily computer-controlled. (2019 prediction)
Can any of you tell me what sites/researchers/corporations/speakers I can follow to keep up with this technology? I can’t seem to find much about it.
by Pauldenice
I just joined this group, although I’ve been reading and writing lots of papers on AI since the mid 1970s.
Of all these readings one really hit a very important aspect in comparing Human versus Artificial intelligence. It was in Isaac Asimov
Robots and Empire
Balantine Books, NY 1985
Page 54
In the whole series, appears one human detective “Elijah Bayley and two high performing humanoid robots who serve as support to the detective…
The two robots in this fiction book are named Giskard and Daneel.
Giskard who is trying to understand human thinking, tells Danneel “Human beings have ways of thinking about human beings that we have not. Giskard’s is searching fo the “laws of humanics” which he is assuming to regulate human thinking just as the famous Asimov’s three laws of robotics regulate completely “robots thinking’ and actions. http://www.auburn.edu/~vestmon/robotics.html
For that, Giskard says he had searched whole galaxiy libraries, trying to discover if such laws governing human behaviour ever existed or if they could be deducted from past human behaviour analysis.
Giskard continues “Every generalisation that I try to make, however broad and simple, has its numerous exceptions. Yet if such laws existed and if I could find them, I could understand human beings and be more confident that I am obeying the Three laws in better fashion.
Giskard keeps on going “Since detective Elijah understood human beings, he must have have had some knowledge of the laws of Humanics.”
Daneel answers: “Presumably. But he knew through something that human beings call intuition, a word I don’t understand, signifying a concept I know nothing of. Presumably it lies beyond reason at my command.”
Giskard again: “That and [robot Memory!] Memory that doesn’t work after human fashion, of course. It lacked the imperfect recall, the fuzziness, the addition and subtractions dictated by wishfull thinking and self interest, to say nothing of the lingerings and lacunae and backtracking that can turn memory into hour-long day-dreaming. it was robotic memory ticking off the events exactly as they had happened, but in vastly hastened fashion. The seconds reeled off in nanoseconds…
Recently a French Neurologist published a fascinating book with a similar view: L’esprit faux I : Mundus est fabula by Benoit Kullmann (http://www.amazon.fr/Lesprit-faux-Mundus-est-fabula/dp/2953532315/ref=sr_1_1?s=books&ie=UTF8&qid=1330203960&sr=1-1)
Human brain’s remarkable capabilities may be the results of our brains imperfections, and neuro sciences mays have started on a wrong path assuming that human intelligence was a perfectly logical system.
Hence my questions:
How can imperfection be embedded in artificial intelligence and will it be done?
Last question, when will computer based intelligences include one of human brains most fabulous possibility: having a “theory of Mind” : How infuriating is it to get a computer program responding besides what one was expecting it to do : “Why didn’t you stupid machine guess what I wanted to do, rather than just executing my commands literally… ”
Note that part of my personal interest in AI was linked to trying to understand the autism way of analyzing received stimuli, and responding to them in a socially acceptable mode.
Fascinating also are brain performances of individuals with a rare condition called “Savant Syndrome” withe truly exceptional capabilities in some domains while having very low cognitive abilities on the average?
Makes one think we may have missed something important in our approaches of brain functions.
Sorry to get somewhat out of the direct subject or singularity, but it seemed to me that this level of machine intelligence may be a fundamental step to achieve: using Human brains imperfections to deliver truly intelligent thinking, and getting this intelligent thinking to build a theory of mind about the people it will interact with.
Feel free to contact me about this comment.
Paul.
by Editor
Paul: interesting ideas. This recent arXiv paper seems to support your comments on the role of imperfections in intelligent thinking: arxiv.org/abs/1202.4495: Stochastic-Based Pattern Recognition Analysis.
It is summarized here: Stochastic Pattern Recognition Dramatically Outperforms Conventional Techniques, http://www.technologyreview.com/blog/arxiv/27598/
by Pauldenice
Thank you very much for your kind comment.
In my latest personal researches (I don’t belong to any university research center) I have considered how imperfections, of all natures, brain or body, may have been at the origin of remarkable creativity, which contributed to lead humanity to the scientific and technology advanced state we are in, at least in OECD countries, so far, but other countries of the world are rapidly coming close, as far as research and technology is concerned…
In a thought experiment I went as far back as the Paleolithic times : Several fossils have proved that human beings had enough empathy to care for elderly or wounded members of their tribes. In order for the tribe to survive with this additional burden, they had to invent new solutions to transport these injured tribes members and also find ways to keep feeding the whole tribe with one less hunter. Creativity may also have come from the very different view of the world that an injured person has, being immobilized for longer periods with an actual different physical point of view on things, and having to find their own strategies to survive.
Over the past 20 years, I have studied at length the exceptional creativity coming from a subgroup of people with autism whose inventive art, often corresponds to Margaret Boden’s “hard creativity definition” :
[It is when] “the world has turned out differently not just from the way we thought it would, but even from the way we thought it could”
M. A. Boden, “The Creative Mind, Myths and Mechanisms”, Basic Book, 1992
Margaret Boden is also one of the top philosopher of sciences who published : “Artificial Intelligence and Natural Man”, Open University Set Book, 1977 Note the date!!!
The “creativity question seems to me to be an essential part of the singularity emergence. Machines will have not only to reproduce and fix themselves, they will have to invent better machines. Or will it be there too a Darwinian process, only more adaptive machines evolutions will survive while the rest won’t .
Consciousness, is another faculty, so far only attributed to human beings and in some rare cases to animals, that may be necessary to reach the singularity level.
However describing and defining “consciousness” remains a big challenge, as very well illustrated in Nicholas Humphrey recent book “Soul Dust”. The Magic of Consciousness”. He concludes that consciousness (however how defined) was most likely a favorable trait in the Darwinian evolution of human beings.
Will consciousness be a necessary evolutionary trait in machines intelligence, in order to thrive, before and after the singularity…
Paul
by sblack
PDF download tested and working. Please contact us at info@Kurzweilai.net if you have further trouble and we can e-mail you the file. THANKS! — Managing Editor
by VictorH
Could you put the pdf file to download again please? I’m trying but it doesn’t open.
Thanks!
by hilander
I think that robots will be able to have intuition.There are 100 or more facial movements that attract our human brain to know what someone is feeling even when they are saying something different.
As Ray says patterns are the key.
I think human emotion is not well hidden in the face, eyes or other signals the body gives off. Lie detector machines can tell if a human is lying. If we can program the very minute changes in facial movements and allow the AI to interpret them into a human emotion I am sure what is called intuition will follow. If they are taught to learn from experience then I just think that we need to be really able to control their growing identities as they become
self aware. It may come a time when some are good and some have learned bad ideas just like today’s children depending on the environment they have been subjected to. That is when we have to think if turning off a badly programmed robot is murder, or if the robot we order around deserves freedom. I guess we should determine how much independent learning we actually want these AI’s to have.
On the other subject I am really concerned that all this wonderful research is going to get further behind if we have a wall street-dollar meltdown. I am a 56 year old single woman, I really would like to see more of this technology advancing here in the U.S. before I die. But with China hiding gold, hacking our weapons systems and conspiring with Russia.
I am afraid like the hydron collider, that no one here thought was important enough to even consider, other countries are going to realize the importance of this research and the funding will come out of Belgium or wherever all those stocks went. I think a global finance system is eminent, and again I would love to know what that means for us with the dollar, and no gold to back it up. Sorry lots of different subjects. I did not notice this thread was so outdated until I posted this sorry
by Editor
Ray Kurzweil’s prediction in The Age of Spiritual Machines (written in the 1990s) that three-dimensional chips are commonly used by 2009 has been further confirmed (within a margin of error) by IBM’s announcement (www.kurzweilai.net/first-commercial-3-d-chip-making-capability-announced) on Dec. 1 , 2011 that Micron will begin production of a new memory device built using the first commercial CMOS manufacturing technology to employ 3-D through-silicon vias (TSVs).