Ray Kurzweil:
"The Singularity is Near" (2005)

(Copyright © 2014 Piero Scaruffi | Terms of use )
"The Singularity is Near" (2005) is the book that popularized the concept of the "singularity". Obviously i have my own (wildly different) ideas (See my book "Intelligence is not Artificial").

Just in the first chapter:

  • "By the end of this century the nonbiological portion of our intelligence will be trillions of trillions of times more powerful than unaided human intelligence": statements like this one mean absolutely nothing. "More powerful" in which way? About what? Will it dribble trillions of trillions of times better than Messi? Will it cook dishes trillions of trillions of times more delicious than my cousin Gisella's? Will it run trillions of trillions of times faster than Usain Bolt? What is the "power" that this sentence talks about? If it is computational speed (which we don't normally call "intelligence" otherwise my pocket calculator of the 1970s was already more intelligent than me), then Kurzweil missed the news that, just about in 2005 when his book was published, Moore's Law stopped working (Intel and the other chip manufacturers started using multiple cores in order to speed up processors but at the cost of creating huge cooling problems and in 2016 William Holt of Intel formally announced that Moore's Law is dead).
  • "There will be no distinction between human and machine". Ever heard of the "extended phenotype"? I guess no. Even before Richard Dawkins published "The Extended Phenotype" (1982), biologists have always known that you cannot separate an animal from its technology: the spider cannot exist without the spiderweb and bees cannot exist without the beehive. Some of Kurzweil's statements are just trivial. Except that, when you replace the humbl spiderweb with Hollywood-style robots, people pay attention.
  • "The future is widely misunderstood. Our forebears expected it to be pretty much like the present". The statement is not only false (many "forebears" expected it to be worse than the present) but it, ironically, applies to Kurzweil himself: Kurzweil thinks that "exponential progress" (the progress that he sees in his lifetime) will last forever. A Roman citizen of the 2nd century expected peace and prosperity to last forever. It turned out that Europe had to wait more than 1,000 years to return to the level of prosperity of the 5th century and 1,600 years to return to that level of peace (actually even longer, because the southern part of the Roman empire, i.e. northern Africa and the Middle East, are not exactly at peace like they were back then). Kurzweil is right that every generation makes the mistake of thinking that a) history ended with me; b) the future will be exactly like the present. What he doesn't seem to realize is that he himself is making the same assumptions. He criticizes the "intuitive linear" view of history which is, in fact, the foundation of his own book: in the future we will simply see more of the present (e.g. the thing that he calls "exponential progress"). He writes that "Almost everyone i meet has a linear view of the future", but he didn't realize that he is one of them (psychologists have a term for this phenomenon but i can't recall the technical term).
  • "Exponential growth is a feature of any evolutionary process". This is obviously false, unless he twists the definition of "evolutionary process". Many evolutionary processes stopped, went backward or simply died out.
  • A diagram titled "Canonical Milestones" is meant to show that "the attributes that are growing exponentially are order and complexity". Go tell the Romans of 2,000 years ago; or, for that matter, go tell the dinosaurs. The diagram does not include the beings truly grew exponentially and that still rule the planet: bacteria. They are not included in Kurzweil's diagram because they hardly prove the value of an exponential increase in complexity since they have virtually no complexity at all. Whether the planet is moving towards more and more complex beings like us or simpler and simpler beings like bacteria only time will tell. So far the bacteria seem to be winning.
  • Kurzweil thinks that machines could "come up with technology to become even more intelligent". Maybe. So far no beings has been able to improve its intelligence. We are as intelligent (or dumb) as the Romans and as the Sumers and as the prehistoric people. Tools have allowed us to do amazing things, and the ability to make tools is certainly one of our great skills; but they have not made us more intelligent (in fact, my grandfather thought otherwise of me, and i think otherwise of the young generations). And certainly tools have not made (of their own intent) more intelligent tools: we made them more intelligent (at the risk of making ourselves dumber - witness people who can't even find their way home without a navigation device).
  • Then there's a long list of trivial statements that mean absolutely nothing like "The rate of paradigm shift is accelerating right now doubling every decade". Information technology undergoes "exponential growth in the rate of exponential growth". Etc.
  • "The traditional strength of machine intelligence include the ability to remember billions of facts precisely and recall them instantly". That is actually a more realistic definition of what "intelligent" machines can do, but this is hardly impressive. I am more impressed by a machine called "electrical switch": you press it and it lights the room. Wow. Not to mention the machine deployed in millions of kitchens that washes your dishes.
  • Alas, and not surprisingly, all of his predictions look terribly out of touch just ten years later. In the first chapter he predicts: "drugs that will allow us to eat as much as we can without gaining weight... will be available in five to ten years". Ten years later such drugs do not exist and they are not even planned. I can easily make a better prediction: they will NEVER exist. You have to change the definition of "eating" to make that prediction work (today or tomorrow or 100 years from now). Then he predicts that "by the end of the decade" (i assume this means 2009) "we will have the requisite hardware to emulate human intelligence with supercomputers". Note that he doesn't say that "we will emulate..." He only predicts that we will have the hardware to do so. Well, it's 2016 and i don't see this hardware, but of course he could claim that we have the hardware and we are too dumb to realize it. "We can expect computers to pass the Turing Test by the end of the 2020s". Of course this depends on what one means by "Turing Test". Computers can easily pass a Turing Test if they are asked trivial questions and/or their human opponent is an idiot. But i can easily predict that no computer will pass a serious Turing Test any time soon, not even in a century, and by "a serious Turing Test" i mean the kind of conversation (about soccer, women, politics, cats, whatever) that we casually engage at the cafe with friends. "Tomorrow's molecular circuits will be based on devices such as nanotubes": a typical Kurzweil prediction, so vague and imprecise that you cannot disprove it, although intuitively it is disputed (i'll let the specialists discuss what progress there has been in applications of carbon nanotubes). It says "tomorrow" (2006? 2016? 3016? 90,345?) and it says "such as": he will claim that his prediction was correct if "tomorrow's" molecular circuits are made of anything other than silicon. "Our experience will increasingly take place in virtual environments": i thought it already did (television, cinema, videogames). Whatever evolution we have next in entertainment he can claim that he was right.
  • Other "predictions" are not predictions but simply summaries of popular science-fiction novellas: nanobots will "reverse human aging", "the entire universe will become saturated with our intelligence", etc. The difference between a humble reader of science fiction and Kurzweil is that Kurzweil takes what he reads in science-fiction books and turns it into a prediction that it will indeed happen (this is another mental disease for which i am sure that psychologists have a technical term).
  • And, for the record, McLuhan never wrote "First we build the tools, then they build us" that Kurzweil quotes in the section "The Six Epochs". The correct sentence is "We become what we behold. We shape our tools and then our tools shape us" and it was written by John Culkin.
The second chapter introduces the "law of accelerating return". We learn that "the Singularity is the inexorable result of the law of accelerating return". Kurzweil defines complexity as "meaningful, nonrandom, but unpredictable information". Then he draws a diagram of events (the abovesaid "Canonical Milestones") that represent an increase in complexity, and identifies an "unmistakable exponential trend". The trend is a fact on that diagram, except of course that he cherry-picked the "meaningful, nonrandom, but unpredictable" events plotted on that diagram. A dinosaur would have picked different events and reached a different conclusion. Then he gives his definition of "order" ("order is information that fits a purpose", whatever that means) and argues that "biological evolution... is no longer leading the pace in terms of improving order" (i assume that there is a "on this planet" missing in this sentence). This is a key assumption here, repeated over and over again throughout the book. It is best formulated later: "The evolution of technology is far faster than the relatively slow evolutionary process that gives rise to a technology-creating species in the first place". He never quite proves this, taking for granted that building the iPhone is as complex or even more complex than creating an eye. I personally think that nothing creating by the human race is even remotely as sophisticated as the eye, let alone the brain. How long nature took to build these things depends on when you want to start counting. If you start counting from the big bang, it took billions of years. If you start counting from the first eye, it took millions of years. How long it took to build the iPhone also depends on when you start counting: if you start counting from the big bang, it took billions of years; if you start counting from the first smartphone, it took only a few years.

The third chapter is an interesting survey of high-tech experiments carried out around 2004. Kurzweil mentions a Stanford Report's 2004 article titled "Scientists Create First Silicon Circuit Using Carbon Nanotubes" about Jeffrey Bokor of UC-Berkeley and Hongjie Dai of Stanford building an electronic circuit made of nanotubes (in 2013 a Stanford team led by Subhasish Mitra and Philip Wong built the first carbon nanotube computer... unfortunately, this computer was made of fewer than 200 transistors and it only offered a clock speed of 1 kilohertz, one million times slower than the processors used in cheap laptops of that year). He mentions a 2004 article in the New Scientist titled "Speed of light may have changed recently " reporting that Lamoreaux and Torgerson of the Los Alamos National Laboratory had found a change in the speed of light over billions of years. We learn that in 2003 Todd Brun at the Institute for Advanced Study studied the possibility of computer sending "information (such as the result of calculations) into their own past light cones". Brun's paper "Computers with Closed Timelike Curves Can Solve Hard Problems" is indeed an interesting read, but ten years later his webpage does not mention a single paper devoted to the topic after that one. Alas, this interesting snapshot of the scientific world in 2004 sounds a bit depressing when read ten years later: there has been very little progress or no progress at all. Eventually, you also understand Kurzweil's trick: pick very recent news (and only very recent ones) that have not yet been tested (and possibly debunked), and build a book around them. No matter how many times you repeat "highly speculative" people will read them as facts. His sentence "These exploratory results will need to be carefully verified" does not undo the fact that he spent three pages discussing them.

The bigger problem with this chapter is the usual one: Kurzweil engages in astronomical statements such as "By 2050 one thusand dollars of computing will exceed the processing power of all human brains on Earth": what exactly is the "processing power" of a human brain? We rarely use the brain to perform calculations, so, if limited to calculations, that statement has been true for a while even before 2004. On the other hand, any human brain can easily engage in a lengthy discussion about a sport event or a political election: is that considered "processing power" or something else by Kurzweil? Whether sheer "processing power" can yield a lengthy political discussion: most people who discuss politics do not have a huge "processing power" (just look at whom they elect) but they can discuss it forever. And a portable computer will "be able to perform the equivalent of all human thought over the last ten thousand years in ten microseconds": meaning what? That this computer will write all of the poems and all of the philosophical essays of the last ten thousand years?

Other statements are just based on bad science, e.g. "10 to the 42nd power calculations per second could be achieved without producing significant heat". Intel begs to disagree, even ten years later. When he talks about "the intelligence created per year will be about one billion times more powerful than all human intelligence today" based on the number of calculations per seconds that machines will be able to perform, he is reducing "intelligence" to "processing power". We will certainly be able to stream more videos than ever, and that will account for the vast majority of traffic on the Internet but only Kurzweil can think that exponentially increasing video streaming is a sign of exponentially increasing machine intelligence. It is not surprising that he reaches the conclusion that the amount of information contained in a rock and in a human brain is roughly the same. If you count the number of atoms, that is obviously true. He is the only one to draw the conclusion that rocks are pretty much as intelligent as human brains. I guess he thinks that elephants are much more intelligent than humans, since their brain contains a lot more neurons. Hence it is easy for him to reach the conclusion that tiny future machines "will be trillion of trillions of times more capable than contemporary biological human intelligence". But this is hardly based on science: it is based on a fundamental misunderstanding of science.

Chapter 4, after a lengthy and somewhat useful summary of progress in neuroscience (although he misses more than 50% of the events that i list in my timeline of neuroscience), Kurzweil can't resist talking about "the accelerating pace of reverse engineering the brain". In fact, we learned very little about the brain until the 1980s when new equipment allowed neuroscience to become truly a science. Since then there has been progress (finally) but hardly exponential: at best linear, at worst painful and tentative. Very few theories of the brain heralded in 2004 stood up to the test of time. Ten years later there are wildly different ideas about dreams, memory, emotions, language and consciousness. Kurzweil might shout "See? That's progress!" Well, let's wait ten years and see if these new theories are any more durable then the ones of ten years ago. It is certainly true that the number of announcements has increased "exponentially" in recent years, but most of those announcements are due to more attention by the media to neuroscience (which may or may not be a good thing) and not to more certainties in neuroscience.

Chapter 5 begins with a statement that i actually endorse: that the 21st century will be shaped by genetics, nanotech and robots (personally, i believe more in software robots than hardware ones). Here Kurzweil confesses his dream: "to slow down disease and aging processes to the point that baby boomers like myself can remain in good health until the full blossoming of the biotechnology revolution, which will itself be a bridge to the nanotechnology revolution". Good luck. Jeanne Clament lived to the age of 122 and died in 1997. Twenty years later nobody has been able to live longer. Kurzweil heroically claims that his "biological age" was 40 in 2004 when he was 56, thanks to "250 supplements (pills) a day". Kurzweil mentions Aubrey de Grey, a prophet of immortality who would write "Ending Aging" (2007) a few years later, introducing him as "a scientist in the department of genetics at Cambridge University" (in reality, a student who graduated in 2000 and has hardly produced anything other than articles and books (perhaps very interesting, but not laboratory research). Kurzweil then begins a long list of new biotechnologies that he obviously expected to stage dramatic progress: i don't think that most of them progressed at all. To corroborate his belief that the pace of progress is accelerating, there is the usual list of articles published in 2003 and 2004 (e.g. a study on self-replicating nanomachines by Tihamer Toth-Fejel for NASA, and a paper by Bernard Yurke of Bell Labs on the first DNA-based nanorobot, neither of which seems to have had follow-ups). Of all the technologies that he mentions two have indeed staged some progress: nanobots and neural nets. There has certainly been progress in designing and building nanobots that can travel into the bloodstream. See my article on biotech. And neural nets are routinely employed in image-recognition problems and even beat weichi masters. The rest of the chapter is dozens of pages of almost delirious speculations taking "recent" studies and blowing them out of proportion and out of context.

Chapter six starts out much better: "Our species has already augmented our natural lifespan through our technology: drugs, supplements, replacement parts for virtually all bodily systems, and many other interventions. We have devices to replace our hips, knees, shoulders, elbows, wrists, jaws, teeth, skin, arteries, veins, heart valves, arms, legs, feet, fingers, and toes, and systems to replace more complex organs (for example, our hearts) are beginning to be introduced. As we learn the operating principles of the human body and brain, we will soon be in a position to design vastly superior systems that will last longer and perform better, without susceptibility to breakdown, disease, and aging." Or: "In human body version 2.0 hormones and related substances (to the extent that we still need them) will be delivered via nanobots, controlled by intelligent biofeedback systems". That was already happening when Kurzweil wrote the book, although progress is not as rapid as he hoped. He also mentions the Internet of Things and virtual reality, that admittedly were under-reported at the time.

Again, the "recent" studies that he mentions don't really hold water. The most important research mentioned in this chapter is arguably Peter Fromherz's research at the Max Planck Institute in Munich. The electrical interfacing of semiconductors and neurons is not trivial because neurons communicate using ions whereas semiconductors use electrons. In 1991 he solved the problem of sensing the electrical field of a neuron on an electronic chip, and in 1995 he solved the problem of stimulating a neuron with an electronic chip (he used neurons of leeches). In 2001 he was therefore able to build a hybrid circuit of electronics and neurons (a snail's neurons). More than ten years later very little progress has been made in building a hybrid electronic-neural circuit with human neurons: you can do it but you'll probably turn into a vegetable. That's the real pace of progress, far from exponential.

And sometimes it is obvious what his personal desires are for this life. He is obsessed with the idea of being able to eat what he likes. For example, he mentions research by Ronald Kahn at the Joslin Diabetes Center and claims that his "Insulin Receptor Knockout Mice" (the title of Kahn's 2003 paper) "demonstrated that the animals were able to eat without restriction yet remain lean and healthy": in fact, ten years later the verdict is still out. But it is funny that, of all problems that afflict the human race, Kurzweil is so determined to make sure that we can eat anything we like as much as we like. Less funny is the fact that Kurzweil has been suffering from type-2 diabetes since he was a young man, and is therefore very interested in news about devices to help diabetes patients. He mentions a project to build an "artificial pancreas" at the Lawrence Livermore National Laboratory. It wasn't an artificial pancreas, but in 2016 the Food and Drug Administration indeed approved a device called MiniMed 670G (alas, only for patients with type-1 diabetes) that delivers insulin 24 hours a day based on the needs of the individual who wears it. It will be marketed by the multinational corporation Medtronic, the world's largest medical technology company (based in Minnesota, not California), that pioneered battery-powered pacemakers. Nothing wrong with this, and we wish the medical technology industry success in finding a cure for diabetes, but it is hardly "exponential progress" and hardly proof that the Singularity is coming: just another device that helps sick people, not really a "cure". Later he mentioned that medicine will eventually allow us to improve/replace so many organs of our body except "the skeleton, skin, sex organs, sensory organs, mouth and upper esophagus, and brain". Then he proceeds to explain how technological progress will also enable us to replace each of these except... the sex organs. So don't worry: they will replace your heart and your brain, but they won't touch your penis or your vagina.

All of these transformations will take place by the 2020s. Then in the 2030s we'll have "Human Body Version 3.0", a body that will incorporate nanotech fabrication "so we'll be able to rapidly alter our physical manifestation at will."

There must be a typo in this chapter where he writes "within another century we will multiply our intelligence by trillions of trillions". So far this book has tried to prove that machines will become trillions of times more intelligent. What will happen to human intelligence has not been discussed. In fact, i am not even sure that today we are more intelligent than prehistoric humans (as usual, it's all about definitions: can you survive in the savannah among lions and hyenas?)

Then the chapter somehow takes a cosmic detour. Kurzweil discusses the "Fermi Paradox", the fact that theoretically there should be millions of intelligent forms of life in the universe but we haven't heard from anybody yet. Kurzweil argues that we are indeed the most advanced technological civilization. I think here he misses a golden opportunity. His theory is that we staged exponential progress in just two centuries and within one more century we'll be in the Singularity: that's a grand total of three centuries, i.e. an ephemeral instant by the cosmic scale. If all civilizations jump suddenly from our state of affairs to the Singularity, it is very unlikely that two pre-singular civilizations can talk to each other. Once we get to the Singularity, if his theory is correct, we'll be flooded with messages coming from all the other civilizations who already got there. Anybody else may simply still in the pre-modern era, unable to send or read radio messages. Instead of this rational explanation that follows from his own assumptions, Kurzweil prefers to come up with a dogmatic "we are alone".

He then delves into the ever popular discussions on the speed of light (in case you haven't guessed already, "ultrahigh levels of intelligence will expand outward at the speed of light"), wormholes, the multiverse, the Hawking radiation, etc. to end with a few naive meditations on consciousness. Needless to say, there are dozens of great books written on contemporary cosmology and consciousness studies.

The last two chapters are Kurzweil's responses to assorted criticism. Surprisingly, these two chapters hold up much better to scrutiny. Then, again, he cherry-picks his critics, so he is just responding to the most superficial and biased critiques. But here he may be in his natural element. He is a bit clownish when he makes prophecies and predictions, whereas he can be razor-sharp rational when he dissects other people's statements.

Along the way Kurzweil repeatedly quotes fellow futurists like Eric Drexler, the prophet of nanotech who wrote the fascinating book "Engines of Creation" (1986), Robert Freitas, author of the fascinating surveys "Nanomedicine" (1999) and "Kinematic Self-Replicating Machines" (2004, with Ralph Merkle), and John Storrs Hall, author of "Nanofuture" (2005). These are all stimulating books, but read them now and you will easily notice how much progress has NOT happened. Nanotech went through its own bubble which then burst, and today most of its promises are still undelivered (the good news for nanotech scientists is that they can still use the same slide presentations that they used in the 1980s).

This book was written by someone who read daily the scientific magazines for a couple of years and collected every sensational announcement, and then organized them into chapters. It is a fascinating snapshot of the science of 2004, but it hardly proves that we live in the age of "accelerating progress". Actually, someone should write a book titled "Where are they now?" surveying all the people mentioned in Kurzweil's book ten or twenty years later. That would indeed be a good way to measure if progress is accelerating (or at least progressing at all).

Kurzweil also scriped the film "The Singularity is Near" (2010). The film is absolutely stimulating, regardless how you feel about A.I., avatars, nanotechnology, etc.

If you are not familiar with Kurzweil's theory of the Singularity, see "The Age of Intelligent Machines" (1990) and "The Age of Spiritual Machines" (1999).

The film opens in 2045. Ramona is an artificial intelligence. Intelligence can be biological or nonbiological. The nonbiological version is not non-human: it is post-human.

Back to our age, the point is made that change is accelerating. Humans just cannot keep up with the pace of change that their technology has caused. Humans need to augment their intelligence. And that's of course also technology. It's technology to keep up with technological change. The ultimate goal is to remove the distinction between technology and "I", Avatars can coexist with humans and help humans enjoy life in a world run by an accelerating incomprehensible technology. As they become more and more realistic and more intelligent, these virtual people can be companions, advisors and even lovers. As an example, we are introduced to Kismet, a robot built in the late 1990s at the MIT by Cynthia Breazeal that feels human. Then Minsky tells us that emotional states are just mechanical states. Then we see the artificial intelligence, Ramona, tell a psychologist how she became what she is: one day a boy (Ray Kurzweil as a child) made a still drawing, and that was her infancy...

Pioneering futurist Alvin Toffler, who wrote "Future Shock" in 1970, says that we need to expand the definition of being human (not what constitutes a human being, but what constitutes humanity).

Then we are taken for a visual tour of our brain via the Allobrain project developed at UC Santa Barbara: "an interactive, stereographic, 3D audio, immersive virtual world constructed from fMRI brain data". Martine Rothblatt tells us that each person will have multiple bodies.

Then the film delves into nanotechnology, presumably to explain one technology that can bring out this marriage of biology and technology, and Eric Drexler, the most famous evangelist of nanoscience, is interviewed.

The film is fair enough to interview a few of the skeptics. The most profound comments come from environmentalist Bill McKibben. His concern is that we are doing something even worse than changing nature: we are changing human nature. The final goal is the death of death (at least for the human species). He is not convinced that he would want to live forever. On the other side of the argument, venture capitalist Vinod Khosla points out that not long ago it was "natural" to die in your thirties but now nobody would sign up for dying at that age. Khosla is convinced that people will always adopt the tools that make them live longer, and even those who say the opposite will do it, given the option.

Back to nanotechnology, Robert Freitas explains how nano-robots (nanobots) can spread into your brain and control every single synapses. They can also replicate so you don't need a big population to start with. As they multiply and evolve, they will de facto provide a continuous upgrade of your brain, just like Microsoft's periodic updates to its operating systems.

James Gashel points out that we don't even need to live in bodies: we will all be alive on the Internet. We won't need a body in order to communicate. The limit will be our imagination, not the material constraints that apply to bodies.

Another skeptic is Bill Joy, who is wary of any chain reaction: once you start it, it is terribly difficult to stop it and the consequences can be catastrophic.

Mitch Kapor is another skeptic. He bet with Kurzweil that no machine will pass the Turing Test before 2029.

Here the film turns into fiction and develops a real storyline. Ramona the artificial intelligence is caught into a political intrigue. An evil human plans to kill her but she is granted a stay of execution and a proper trial. Her defense attorney is the notorious Alan Dershowitz who defended sport star O.J. Simpson at a famous trial. The prosecutor claims that Ramona is nothing but a piece of matter. Dershowitz begs the judge to allow for a Turing Test to decide whether Ramona qualifies as human. The judge accepts. Ramona is tested against three humans. A jury has to decide which of the four is the machine. Ramona is reminded of what a wise man told her: humans feel love in their heart. She replies in a convincing manner and none of the judges guesses which of the four tested is the machine. She passed the Turing Test and her life is spared.

The end is a lengthy monologue by Kurzweil behind colorful visual effects. He talks about how machines will eventually spread human intelligence to fill the whole universe. (Cinematically speaking, this is also the best part of the film - see this snippet).

For what it's worth:

  • I see a lot of regress and not so much progress (read "Regress")
  • I think that in the short term life expectancy will decrease, not increase (read "Revising the Myth of Longevity"). I think Kurzweil's data on longevity are superficial. The gain in life expectancy has mostly come from reducing infant mortality, mainly from lower rates of respiratory diseases such as turberculosis. However, the decline in respiratory diseases predates modern medicine: it started in the 19th century. It would be difficult to find a "technological" cause: better nutrition, and therefore a stronger immune system, might be a more likely cause. It is debatable if today we are moving towards stronger or weaker immune systems. The science of some of his arguments is also superficial. Genes don't make proteins, nor do they make copies of themselves. Genes provide a mere blueprint. What makes proteins and copies of genes is a complex apparatus within the body. If you have the right genes but your body sucks, you get sick no matter what. It is debatable if today we are moving towards stronger or weaker bodies.
  • I believe machines will pass the Turing Test but not because they are becoming more intelligent: they will pass the Turing Test because humans are becoming less intelligent (read "Machine Intelligence and Human Stupidity")
  • The idea of the coming singularity requires an incredible leap of faith. First of all, we are asked to believe that machines are soon going to match the cognitive skills of humans, which is hard to believe given how dumb machines still are compared to even the dumbest human beings. Sorry but, as of today, we don't even have software that can write other software. Secondly, we are asked to believe that machines will achieve a degree of intelligence that is beyond the capability of humans. I have no doubts that there is a limit to what the human brain can do, the so called "cognitive closure", and that there might well be (now, in the future, in the past) beings who have superior capabilities, but i am missing the proof that machines created by our intelligence (or created by machines created by our intelligence) will be able to do something that our intelligence cannot do. Thirdly, these arguments rely on the fallacy of assuming that human cognitive progress (the progress of the human mind) ended with you. Today (2012) that perception is partially justified because technological and scientific progress (the way i see it) has slowed down compared with a century ago. Anybody who has studied the history of Quantum Mechanics and Relativity has been amazed by the incredible (literally incredible) insight that those scientists had, by the leaps of imagination that led them to understand a reality that they could not see. By comparison, today it is easy to fall into the fallacy of believing that the human mind has reached a maximum of creativity and will never improve beyond what it has created so far, i.e. intelligent machines are the terminal point of human cognition. However, it is likely that sooner or later the human mind (for whatever sociopolical, economic or cultural reason) will resume its progress and then the next step in creativity might well take us in a different direction and lead us to invent things (not machines) of a different kind that will make intelligent machines look as obsolete as windmills look today. Today's electronic machines may continue to exist and evolve, just like windmills existed and evolved and did a much better job than humans at what they were designed to do, and machines might even build other machines that will do that job better; but we are assuming that, in the meantime, humans will stop being creative actors and will become merely passive spectators.
While i was listening to all these gurus predicting this spectacular future of eternal avatars, the batteries of my smartphone died. I had recharged them yesterday night. Technology reminded me how far we are from all of this. It's fine to live forever but for now can someone please give me batteries that will last at least one day, just one day?

See also: my book on the Singularity.

TM, ®, Copyright © 2014 Piero Scaruffi