- Paperback: 672 pages
- Publisher: Penguin Books (September 26, 2006)
- Language: English
- ISBN-10: 0143037889
- ISBN-13: 978-0739466261
- Product Dimensions: 6 x 1.4 x 9.2 inches
- Shipping Weight: 1.5 pounds (View shipping rates and policies)
- Average Customer Review: 4.1 out of 5 stars See all reviews (426 customer reviews)
- Amazon Best Sellers Rank: #8,204 in Books (See Top 100 in Books)
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
The Singularity Is Near: When Humans Transcend Biology
Use the Amazon App to scan ISBNs and compare prices.
Frequently bought together
Customers who bought this item also bought
From Publishers Weekly
Starred Review. Renowned inventor Kurzweil (The Age of Spiritual Machines) may be technology's most credibly hyperbolic optimist. Elsewhere he has argued that eliminating fat intake can prevent cancer; here, his quarry is the future of consciousness and intelligence. Humankind, it runs, is at the threshold of an epoch ("the singularity," a reference to the theoretical limitlessness of exponential expansion) that will see the merging of our biology with the staggering achievements of "GNR" (genetics, nanotechnology and robotics) to create a species of unrecognizably high intelligence, durability, comprehension, memory and so on. The word "unrecognizable" is not chosen lightly: wherever this is heading, it won't look like us. Kurzweil's argument is necessarily twofold: it's not enough to argue that there are virtually no constraints on our capacity; he must also convince readers that such developments are desirable. In essence, he conflates the wholesale transformation of the species with "immortality," for which read a repeal of human limit. In less capable hands, this phantasmagoria of speculative extrapolation, which incorporates a bewildering variety of charts, quotations, playful Socratic dialogues and sidebars, would be easier to dismiss. But Kurzweil is a true scientist—a large-minded one at that—and gives due space both to "the panoply of existential risks" as he sees them and the many presumed lines of attack others might bring to bear. What's arresting isn't the degree to which Kurzweil's heady and bracing vision fails to convince—given the scope of his projections, that's inevitable—but the degree to which it seems downright plausible. (Sept.)
Copyright © Reed Business Information, a division of Reed Elsevier Inc. All rights reserved. --This text refers to an out of print or unavailable edition of this title.
From Bookmarks Magazine
Kurzweil is one of the worlds most respected thinkers and entrepreneurs. Yet the thesis he posits in Singularity is so singular that many readers will be astoundedand perhaps skeptical. Think Blade Runner or Being John Malkovich magnified trillion-fold. Even if one were to embrace his techno-optimism, which he backs up with fascinating details, Kurzweil leaves some important questions relating to politics, economics, and morality unanswered. If machines in our bodies can rebuild cells, for example, why couldnt they be reengineered as weapons? Or think of singularity, notes the New York Times Book Review, as the "Manhattan Project model of pure science without ethical constraints." Kurzweils vision requires technology, which we continue to build. But it also requires mass acceptance and faith.
Copyright © 2004 Phillips & Nelson Media, Inc. --This text refers to an out of print or unavailable edition of this title.
If you are a seller for this product, would you like to suggest updates through seller support?
Top Customer Reviews
Anyone who has ever played around with the arithmetic of compounding and exponential growth knows how crazy the numbers get as growth feeds on itself. The phenomenon is quite real in the world, and it describes everything from viral epidemics to Warren Buffet's fortune. Kurzweil applies the exponential growth paradigm to the future of technology. He sees not only change itself accelerating, but the rate of change too, if you can go back to your high school calculus and wrap your mind around that stomach-churning concept. The math starts quickly approaching infinity, which is why it's so weird.
"Singularity" is a common term-of-art among theoretical physicists, who apply it to a variety of seemingly irrational constructs, such as an infinitely large mass compressed towards an infinitely small point. Kurzweil co-opts the term for his own purpose here to mean the point in time where artificial intelligence starts exceeding human intelligence. Thereafter, it takes over its own programming and, being so powerful, does a better and better job of it. Because things are already moving so fast today, the accelerating rate of change means that Kurzweil's Singularity is closer than even optimists might imagine - hence the book's title. He projects it to occur somewhere in the middle of this century. Afterwards, nothing will ever again be the same.
In physics, unimaginable things start happening at singularity points, like energy explosions within black holes. Following Kurzweil's Singularity, the most garish science fiction fantasies start becoming commonplace. The combination of genetics, nanotechnology and robotics - which he refers to collectively as GNR - will transform all aspects of human existence. He believes, for example, that nanobots released into a person's bloodstream, will facilitate a comprehensive (that is to say, 100%) map of that person, including genetic code and nervous system, that can be uploaded and downloaded at will onto new "substrates". In other words, robotic copies of human beings - body, mind, memories, and (one presumes) soul - can be made that will appear indistinguishable from the originals. And for that matter, those originals themselves can be re-shaped at will, giving us all the opportunity to become brilliant, strong, happy, and beautiful.
Kurzweil tells us that artificial circuits replicating themselves at a molecular level will merge with the biological circuits that constitute our nervous systems, giving rise an "enhanced" human super-intelligence. Once this starts happening, what we now call the Internet will in effect become telepathic, giving these enhanced humans instantaneous access to all available knowledge and information as they fashion their brave new world. You see how explosive this gets? And it's just the beginning.
Once the process gets underway, the evolving super-intelligence keeps expanding until it permeates the entire planet and, still accelerating, eventually the universe. Kurzweil suggests that movement though time-space "wormholes" should one day facilitate rapid travel beyond our own galaxy, taking the process literally everywhere.
I realize that my amateur's survey of Kurzweil's thinking here makes him sound like a crank. However, let there be no mistake: he is an accomplished scientist and a highly sophisticated thinker. MIT-trained, he's an expert in artificial intelligence and has put his ideas into practice as a successful tech entrepreneur. Most of this book is not even devoted to prognostications, but to an in-depth review of research currently underway that lays the practical groundwork for virtually everything he talks about (except maybe the wormhole business). While he makes numerous leaps of faith in taking us from here to there, none of his forecasts represent sheer fantasy. He is an extremely good writer, and while staying true to what is in fact pretty complex science, describes it all in a way that makes it reasonably clear to lay readers.
For all his hardcore materialism, Kurzweil also has a whimsical streak. Every 50 pages or so, he breaks up his text with imaginary light-hearted debates among himself (appearing as "Ray"), various historical figures - Darwin, Freud, etc. - and a person named "Molly", who seems to be a student. Molly is bright, curious, skeptical, and not in the least bit awed by Ray or the others. The thing about Molly is that she appears in two separate guises: Molly 2004 (the year this book was being written), and Molly 2104, which is of course well beyond the Singularity. One of Kurzweil's key forecasts is that future science will learn how to arrest and even reverse the aging process, allowing people more-or-less to live forever at whatever age they choose. So Molly has made it through the Singularity and returned as a still-young woman to speak about it from experience.
Kurzweil is fully aware of the potential downside to his vision. He devotes one long chapter to what he calls "The Deeply Intertwined Promise and Peril of GNR". He devotes another even longer chapter to responding to critics, who have attacked his ideas from every possible perspective. While he treats most criticisms respectfully, in the end he largely dismisses them all. One partial exception and the one specific fear he himself does seem to harbor is of self-replicating nanobots. He and other scientists who seriously debate such stuff even have a short-hand term for this specter: The Grey Goo Problem. Were self-replication somehow to spin out of control, Kurzweil explains to us that in a matter of days it could, in theory, consume the Earth's entire biomass and reduce it to "grey goo". This is indeed a troubling prospect, since this endangered biomass includes all of us.
Interestingly, the cluster of criticisms that he responds to most gently are those arising from a spiritualist perspective. In one of his imaginary debates with "Molly", she repeatedly asks "Ray" if he believes in God. Ray surprises by dodging the question every time rather than saying no. Badgered into a corner, he finally avers: "For the sake of your question, we can consider God to be the universe, and I said that I believe in the universe." This sounds suspiciously like a yes, albeit with a twist. He then goes on to explain how his entire vision can be described as a picture of the universe "waking up" as enhanced human intelligence pervades its many corners. Religious people of an unorthodox bent might be tempted to embrace this image as God's self-realization. Fundamentalists of every stripe, however, were they to take K's cosmology seriously at all, would view it with disgust as the self-realization of God's Opposite Number.
For me, the most unnerving question that this book triggers is who will control these accelerating technologies. Reading through many passages of the book, I found it hard not hard to be thinking about Nazi scientists beavering away at the design of their Master Race, or North Korean labs re-programming the neural patterns of citizens lacking enthusiasm for Kim Jong-Un. Kurzweil seems to trust in the pragmatic good will of the scientific community, buttressed by regulation. However, not all scientists have good will, and he says nothing about who he supposes will regulate the regulators. I also find it hard to see what joy or challenge there could be in a world where machines or enhanced humans dominate everything. People choosing not to become "enhanced" would either have it forced upon them or face life as a sub-species. The line between utopia and dystopia here is pretty fuzzy, and I find it a little scary that Kurzweil doesn't seem to care. Maybe I've seen too many science fiction movies.
All that aside, I highly recommend this book. Decades ago when I was in college I used to describe about every other book I read as "changing my life", as we said in the day. Nowadays, no book changes my life, although the best ones still move the needle for me. Whether I like it or not, this one has me looking at things a little differently than I did before.
With his last couple of doublings on the exponential curve of human progress, he reaches deep into faith more so than science. There are some major biases and contradictions in his theories, along with getting one of his main premises completely wrong.
He thinks we can create AI when the computing power of a machine reaches the 125 trillion synapses you have in your average brain. From that point on, we never look back and rule the universe as human-machine hybrid gods.
There a couple of really big problems with that assumption.
One, he assumed each synapse in our brains is a logical switch, which at the time he wrote the book was considered a known fact. Fair mistake. But the very progress of technology proved him wrong. The latest in neuroscience has shown that the brain is vastly more complicated than that and each synapse is much closer to a very sophisticated micro-processor with thousands of gates than to a simple logical switch. So, to put that in perspective--a single developed human brain has 125 trillion (or thereabouts) processors working seamlessly in parallel, taking input from 200 billion nerve cells, with extremely low energy requirements and low heat output.
For comparison, the entire human race currently operates several billion micro-processors. I have no doubt that number will increase several-fold in the next couple of decades. The second most powerful supercomputer in the world, the Cray Titan, has 37 thousand. The next generation will have less nodes, albeit faster ones. Yes, they will get smaller, faster, more efficient. However, to think that we'll replicate in machine form the true networked complexity of the human brain in the next 10 or 20 years is just nuts, given that it takes 3-5 years to actually build the next generation of supercomputers and even more years before people actually figure out what to do with them. So, on the ‘hardware’ requirements for AI, under his original assumption, I think he was completely off and by orders of magnitude.
The even bigger issue is on the software side—his premise is that we only need to simulate the output of a super-complex system like the brain, not necessarily know how exactly the black box works. Hence, once we understand how the human brain works, we don’t need to rebuild it in machine form, we only need to build something which given the same inputs provides the same output. Then you’ll have true AI because we’ll have a machine that can self-learn, write unconstrained code, etc.
The problem here is that he has a very fundamental chicken and egg problem. The issue is that AI and digitalizing the brain is the very cornerstone of everything else in his writing—including nanobots (which even he admits will have their software written by a higher form of intelligence, not human) which truly kicks off the next stages of human-machine evolution. Right now we cannot even map what the brain really does on a molecular level, let alone accurately replicate it.
Furthermore, I know some very bright people who code for a living and even the best and brightest struggle with limited parallel programming and it gets exponentially more difficult with each level of parallelism you introduce. I am yet to meet a coder, no matter how self-confident, who will, with a straight face, claim that they can write software that can replicate or even approximate 125 trillion processors working in parallel. Kurzweil says that an AI unconstrained by physics or biology can write such code. So, while Kurzweil assumes that with increased computing power we’ll just figure such a true AI out, I’m not so sure it is that simple.
So, how do we get AI to begin with, if it truly takes a higher order of intelligence to write such an AI?
Overall, just like many futurists, there are hits and misses. I believe that when it comes to machines becoming pervasively widespread in everything we do, he is right on and has been for two decades. The next part, where humans become machine gods borders on religious (technology being the god here) fanaticism more so than on sound science.
Anyhow, time will tell, but I think he is at the very least off (if not completely wrong) on his timing—and by a lot.