Comprehensive coverage

The approaching merger between mind and machine / Ray Kurzweil

The accelerating pace of technological progress dictates that our intelligent creations are expected to eclipse us soon - and that their creations will eventually eclipse them as well. Futurist Ray Kurzweil predicts that by the middle of the century we will no longer be able to distinguish man from machine.

Ray Kurzweil in a lecture at Stanford University in 2006. From Wikipedia
Ray Kurzweil in a lecture at Stanford University in 2006. From Wikipedia

The accelerating pace of technological progress dictates that our intelligent creations are expected to eclipse us soon - and that their creations will eventually eclipse them as well. Futurist Ray Kurzweil predicts that by the middle of the century we will no longer be able to distinguish man from machine.
At some point, in the first part of the 21st century, the intelligence of machines will catch up with the intelligence of humans. Within a quarter of a century, there will be machines that at first glance appear to have human intelligence, emotions and skills of all kinds: from music and other creative skills to physical movement. They will claim that they feel emotion, and unlike today's virtual characters, their claim will sound very convincing. Around 2020, the processing power of a $1,000 computer will equal that of a human brain. In 2029, the software necessary for full intelligence will be almost completely completed and the average personal computer will be equal to about 1,000 brains.
Once computers achieve this level of intelligence compared to humans, they will inevitably wander forward. For example, if I am learning French, I cannot easily load this learning into your mind. Because our learning process involves building sequences of incredibly complicated link patterns between the nerve cells in the brain, which also involve the concentrations of biological substances, nerve messengers, that allow signals to pass between the nerve cells. We have no way to quickly transfer these patterns to another brain. But the ability of our non-biological clones to quickly transfer information between them will allow them to instantly share what they've learned with billions of other machines. Eventually, non-biological entities will lack not only all their knowledge but all our knowledge as well.
When this happens, there will no longer be a clear distinction between man and machine. Already today we are implanting computers, or neural implants, directly inside people's brains to dampen the effects of Parkinson's disease and the tremors resulting from multiple sclerosis. We use cochlear implants to restore hearing. In the USA today a retinal implant is being developed that could provide the blind with some type of visual perception by replacing some visual processing circuits in the brain. A team of scientists at Emory University implanted a chip in the head of a stroke paraplegic that allowed him to use his brain to move a cursor on a computer monitor.
In the 20s of the 21st century, neural implants will upgrade our sensory experiences, memory and thinking. In 2030, we will be able to replace a phone conversation with a friend with a virtual meeting, for example in a game reserve in Mozambique, that will look completely real. You can have any kind of experience - business, social or sexual, with any person, real or imagined, regardless of the physical proximity between you.
How life and technology develop
In order to better understand the type of predictions I just presented, it is important to note that information technology advances at an exponential (exponential) rate. An exponential process starts slowly, but eventually it progresses very quickly. (You can read a full documentation of this argument in the books "The Singularity is Close".)
The evolution of biological life and the development of technology both tread the same path: they take a long time to begin, but the stages of progress build on top of each other, until progress erupts in an ever-increasing surge. Right now we are entering this breakout phase in the technological development curve.
Note: It took billions of years for the Earth to form. Another two billion years until life began, and almost the same amount of time was needed for molecules to organize themselves to form the first multicellular plants and animals about 700 million years ago. The pace of evolution accelerated when mammals inherited the land about 65 million years ago. The appearance of the primates meant that the rate of evolutionary progress was measured in only millions of years, until the appearance of modern man about 500,000 years ago.
The technological development is actually the continuation of the evolutionary process that led to our own existence in the first place - the creatures that create technology. It took tens of thousands of years for our ancestors to understand that if they sharpened a stone on both sides they could make useful tools. Then, at the beginning of the previous millennium, the time required to change a technological paradigm shrank to hundreds of years.
The rate of technological progress continued to accelerate in the 19th century and equaled that of the 10 centuries that preceded it. In the first two decades of the 20th century, progress equaled that of the entire 19th century. Today, important technological revolutions occur every few years. For example, the Internet, a means of communication and commerce that is already everywhere, did not exist just 20 years ago. And only about ten years ago almost no one used search engines.
Computer technologies are developing on the same exponential growth path. For the last decades, the important factor in this expansion was described by Moore's Law. Gordon Moore, one of the founders of Intel, noticed in the mid-60s that technologists doubled the density of transistors in integrated circuits every 12 months. That is, computers have doubled in regular cycles both their capacity and speed per unit price. In the mid-70s, Moore updated his observation and set the duration of the doubling cycle, according to a more reliable estimate, at about 24 months, and this trend persisted throughout the years.
After decades of dedicated service, Moore's Law will be forced to retire in about 2019. By then the transistors will reach the width of several atoms. But new computer architectures will continue the exponential growth of computing. For example, calculation cubes that are already being designed will contain thousands of layers of circuits, and not just one as in today's computer chips. Other technologies promise an order-of-magnitude increase in computing density, including circuits based on carbon nanotubes, optical computing, crystal computing and molecular computing.
We can easily see the march of computing forward through a graph depicting the calculation speed (in units of instructions per second) per 1,000 dollars (in constant dollar value) of 49 famous calculating machines from the 20th century. The graph looks like an example taken from a textbook dealing with exponential growth: computer speed per unit price doubled every three years from 1910 to 1950, every two years from 1950 to 1966, and every year today. It took 90 years to get to the first computer capable of executing a million instructions per second (MIPS) at a price of $1,000. Today we add several more MIPS to a $1,000 computer every day.
Credit: Daphne Axel
Why are yields accelerating?
Why do we find that exponential development occurs in the biological world, in technology and in computing? This is the result of a fundamental feature of every evolutionary-developmental process, a phenomenon I call the "Law of Accelerating Returns". When order increases at an exponential rate (a phenomenon that reflects the essence of evolution), the time period between notable growth events shortens. Progress is accelerating. And the returns - the valuable products that development yields - accelerate in a non-linear manner. The increasing growth of computing performance relative to price is one important example of such accelerating returns.
A common criticism of forecasts claims that they rely on an unjustified continuation of current trends (extrapolation), without taking into account the forces that may change these trends. But evolutionary development accelerates because it builds on the achievements of the past, including an improvement in the ability of evolution itself to develop further. The resources that development needs to utilize in order to continue its exponential growth are actually its own increasing order, and the chaos in the environment in which the development process takes place, which provides the possibilities for future diversity. Both of these resources are virtually unlimited.
The Law of Accelerating Returns shows that around 2020 a personal computer priced at $1,000 will have the computing power of a human brain - 20 million billion calculations per second. The estimate is based on areas of the brain whose operation has already been successfully reconstructed in computer simulation. In 2055, computing power at a price of $1,000 will be equal to the computing power of all human minds in the world (and I could be wrong of course by a year or two).
for intelligent programming
This is the prediction regarding the power of calculation, a necessary but not sufficient condition to achieve human-level intelligence. The intelligence software is more important than that.
One approach to creating this software is to meticulously program the laws of complex processes. Another approach is programming using "complexity theory" (also known as chaos theory). In this approach, self-organizing algorithms gradually learn patterns of information in a manner similar to human learning. One such method, neural networks, is based on simplified mathematical models of mammalian neurons. Another method, called genetic (or evolutionary) algorithms, is based on the gradual development of smart solutions in a simulated evolutionary environment.
But eventually we will learn to design intelligence by copying the best intelligent entity we can achieve: the human mind itself. We're going to reverse engineer the human brain, and luckily it's not copyrighted!
The most direct way is a destructive scan: take a frozen brain just before it stops working, slice it into thin slices and examine them, one by one, to discover each nerve cell, each interneuronal connection and measure the concentrations of neurotransmitters in each interval between the nerve cells (called synapses). One executed murderer has already donated his brain and body for such a scan, and all 15 billion of his bits are accessible on the website of the American National Library of Medicine (http://www.nlm.nih.gov/research/visible/visible_gallery.html). The resolution of these scans doesn't come close to our requirements, but the data at least allows us to start thinking about these issues.
We also have non-invasive scanning methods, such as high-resolution magnetic resonance imaging (MRI) and others. State-of-the-art scanning methods are able to show single interneuron connections in a living brain and display them in real time as they send a signal. The improved speed and resolution of these methods will eventually allow us to decipher the connections between neurons. The accelerated improvements are also due to the law of accelerating returns, because powerful computing power is the main factor that enables high resolution imaging.
Another possible approach would be to send microscopic robots (or "nanobots") in the bloodstream to the brain and program them to explore every nerve and monitor the brain's connections and the concentrations of neurotransmitters.
the amazing journey
Although such small sophisticated robots are at least a few decades away in the future, using them to examine the most hidden corners of our bodies would be far-reaching. They will be able to communicate with each other wirelessly and report their findings to external computers. The result will be an image of the brain from the inside, obtained in a non-invasive way.
Most of the technologies needed to realize this scenario already exist, but not yet at the necessary microscopic scale. However, shrinking them down to the necessary tiny sizes will reflect the essence of the law of accelerating returns. For example, transistors in integrated circuits have shrunk by about 5 times, in each of their three dimensions, every 10 years.
But the capabilities of these implanted nanobots won't be limited to passive roles like monitoring. In the end it will be possible to design them for direct communication with the neural circuits in our brains and to strengthen and expand our mental abilities. Already today we have electronic devices that are able to communicate with nerve cells, detect their activity and cause neighboring nerve cells to send a signal or prevent it from being sent. The implanted nanobots will be able to reprogram the neural connections to allow us virtual reality experiences and upgrade various mental skills such as the ability to recognize patterns.
In order to decipher and understand the brain's information processing methods (which happen to combine digital and analog methods), it is not necessary to see every connection and connection, because in every area of ​​the brain there is quite a lot of duplication. Already today we are applying some of the insights we acquired in the first stages of this reverse engineering process. For example, in speech recognition we were able to decode and copy the first stages of sound processing in the brain.
But maybe instead of scanning the brain to understand, it would be more interesting to scan the brain to download it to the computer. We can map the locations and connections of all nerve cells and synapses and all concentrations of nerve messengers. Then, we can recreate this entire organization, which contains the brain's memory, on a digital-analog computer.
For this we will need to understand the local processes in the brain, and progress in this area is already happening. Theodore W. Berger and his colleagues at the University of Southern California have built integrated circuits that precisely match the processing characteristics of clusters of neurons of considerable size. Carver A. Mead and his colleagues at the California Institute of Technology have built various integrated circuits that mimic the digital-analog characteristics of mammalian neural circuits. There are simulations of brain areas responsible for visual processing and of the cerebellum, responsible for building skills.
Drawing a perfect map of the human brain is not as daunting a task as it sounds. The Human Genome Project also seemed impractical when it was first proposed. At the rate at which genetic codes were scanned 20 years ago, it would have taken thousands of years to complete the mapping of the genome. But according to the law of accelerating returns, the ability to sequence DNA doubled every year, and the project was completed in 2003.
In the third decade of this century, we will be able to create complete and detailed maps of all the features relevant to calculation in the human brain and reproduce this structure in advanced neural computers. We can also give our machines a variety of bodies, from virtual bodies in virtual reality to bodies composed of swarms of nanobots to humanoid robots.
The question of consciousness
Such possibilities raise a variety of fascinating questions and issues. Let's say we scan someone's brain and initialize the "brain file" we get with appropriate computing means. Will the resulting entity have consciousness? Apparently this entity will have a personality, history and memory very similar to the original. There are people for whom this is enough to define consciousness. Others, such as the physicist and author James Tarfil, believe that no logical reconstruction can reach human consciousness. However, Trefil accepts the possibility that computers may become conscious in a new way of their own.
When exactly do we accept as a fact that some entity has consciousness, self-awareness and free will? How do we distinguish between a conscious process and a process that behaves as though it is conscious? Is it enough for the entity to tell us very convincingly: "I'm lonely, please give me company"?
If you ask the "person" inside the machine, he will firmly claim that he is the original person. If, for example, we scan me and initialize this information in a neural computer, the person that will be created will think that he is (and has always been) me (or at least will act that way). He will say: "I grew up in the borough of Queens in New York, I went to college at MIT, I stayed to live in the Boston area, I went into the scanner there and woke up in this machine here. Hi! This technology really works."
But wait a minute, is that really me? What is certain is that the old Ray (meaning me) still exists inside my brain made of carbon-based cells.
Will the new being be able to have spiritual experiences? Since the brain processes in the machine will be essentially the same as those of the original human, its behavior will also be similar. So there is no doubt that the machine will claim the full emotional and spiritual experiences that man also claims.
There is no objective test capable of determining absolute consciousness. We have no ability to objectively measure subjective experience (because this is related to the essence of the concepts "objective" and "subjective"). We can only measure things that correspond to it, like behavior. The new entities will appear conscious, and whether they are conscious or not will not affect their behavior. As we debate today the question of the consciousness of non-human beings, such as animals, so we will surely debate the possibility of the consciousness of non-biological intelligent beings. Practically speaking, we had to accept their claims. They will be angry if we don't.
Before the end of this century, according to the law of accelerating returns, the species production technology of the planet, that is us, will merge with our own technology. When this happens, we might ask: What is the difference between a human brain amplified a million times by means of neural implants and a non-biological intelligence based on the reverse engineering of the human brain which was then amplified and expanded?
The mechanism of evolution used the innovation of one era (humans) to create the next (intelligent machines). The next milestone will be when the machines produce their next generation without human intervention.
The evolutionary process accelerates because it builds itself using its own resources to continue evolution. Humans have defeated evolution. We create intelligent beings in significantly shorter periods of time than the time it took for the evolutionary process to build us. Human reason, the product of evolution, has surpassed it. Therefore, the intelligence we now create in computers will soon achieve the intelligence of its creators.

About the author
Ray Kurzweil is the CEO of Kurzweil Technologies. He headed the team that built the first machine to translate printed text into speech, the first system for optical character recognition (OCR) of any font, the first text-to-speech generator, the first music generator capable of producing grand piano sounds, and the first commercial speech recognition system in a wide vocabulary. In 2009 he founded the Singularity University to cultivate leadership for technological innovation.

37 תגובות

  1. A brain computer interface (BCI), often referred to as a brain machine interface (MMI), is a wireless communication pathway between the brain and an external device. BCI is often used to assist or correct cognitive, sensory or motor functions in the person.

    This research began back in the 70s at the University (UCLA) in Los Angeles under a grant from the National Science Foundation. In this preliminary study, they succeeded for the first time in translating the brain's activity into expressions in a computer interface.

  2. Needs:
    It was not a matter of assuming what was requested, but an answer to Yair's question.
    I also assume that there is a high chance that this is also a stage we will have to go through on the way - at first we will "copy" the brain of a living being and try to make it work without fully understanding all the mechanisms and only then we can use it to crack the mechanisms that make it work.
    It is worth mentioning that today there are already machines whose capabilities exceed ours in many fields (such as aviation, progress at sea and on land, carrying journeys, and even calculations).
    The point is that in all these machines - we satisfy the desire. Except for accidents, they don't do anything without us wanting them to.
    There is a chance that the ability to think in a really creative way - also requires the existence of a will and it is clear that we must make sure that it is not a will that contradicts our will.
    Seeing this problem, Asimov formulated the laws of robotics, but the problem is apparently more complicated.

  3. An interesting question about the needs of the machine came up in the discussion.

    Michael assumed that the desires of the machine would be the same as the needs of the human brain upon which that future supercomputer would be based. It's a bit of a presumption - because if the supercomputer can't go beyond the intent of its creators - it obviously won't be a threat.

    The interesting question is what happens if we assume (a necessary assumption in my opinion) that the supercomputer will not be subject to the intention of its creators, and will be able to change their habits even though they were the basis of its creation. So for example, I want to believe about myself that I can change my habits (to a certain extent...) despite hundreds of thousands of years of evolution that instilled these habits in me. There is no reason not to make a similar assumption about an intelligent being with abilities that exceed mine.

    And another problem for Dokhta - what are the needs of a computer system? The system needs energy, raw materials - and if it is human in terms of its essence, also love and empathy. We usually imagine advanced computer systems as the threat of a heartless advanced being - there is also the opposite possibility.

    If the system has abilities that surpass ours - maybe it can also reach better spiritual understandings than ours and treat both the cow and the human better than we do. In the meantime - we have to improve ourselves, so that when the day comes, when the supercomputer is built on our basis - we can hope that it will have a good model to copy from.

  4. Ray says interesting things and also substantiates some of them mathematically, by basic laws.
    Let's not forget that one of the famous people who predicted fateful progress for humanity at an exponential rate was Thomas Malthus, who is considered the father of mathematical predictions, but his predictions failed miserably, because he did not take into account unthinkable phenomena and broken cards.
    Kurzweil himself says that Moore's law will give way - will his laws stand firm.
    I tend to agree with him, but think that he is a little ahead and does not consider the problems of the human race as an existential basis for the development of technology. Survival is first and foremost, and today, however it appears, resources will be directed there.

  5. I think intelligent computers that are able to think just like humans are only a few decades away from us, it's going to be amazing, can't wait for it already.

    : )

  6. Nice as bedtime reading, nothing more.
    At least he doesn't predict a flying car for us, that's good.

  7. And regarding talking about things you don't know:
    I don't like to make claims that I don't know if they are true.
    It turns out that there are those who like it or who really think they know what I think they don't.
    You can talk about speculations and present them as such, but to present this kind of prediction is charlatanism in my eyes.

  8. Year:
    I thought I explained clearly.
    What drives man will drive the machine.
    It is not the need but a feeling in man that evolution created because of the need and in the machine the human being will be created by the one that will duplicate a human brain.
    What is not clear here?
    Read my previous response again: we do nothing because of necessity because we do all things because of impulses that have been ingrained in us by evolution.
    The fact that evolution drowned them out of necessity really doesn't matter.
    What is important is that they are inherent in us and because of them we act.
    They will also be inherent in the computer - not because of evolution but because of us.

  9. Yair
    That motivation of the machine (the truly intelligent, and the first to be created) will be the derivative determined by the person/people who created it.
    As long as such a machine is not created, we will not be able to learn what motivates it to be more intelligent.
    But the first intelligent machine that will be created will have a motivation that will be derived from the programming that will be set for it.

  10. ghosts,
    You are talking about human motivation, which needs no proof, my question is about the motivation of a machine.
    point,
    If the machines execute man's software, they will not pass the level of man's servers. The article predicts machines that will not only replace man, but will also be replaced by more sophisticated machines, seemingly a continuation of natural evolution. And what is missing in this prophecy is what will be the factor that will drive the intelligent potential of the machines.

  11. Yair, the needs (in the language of machines they can be called goals), are nature's primitive ways of instructing the body to do something when the goal that stands beneath all these is survival and culture. There is no problem to write software with any chosen goal. But why occupy the computer with food when it can be programmed to play games and win (the goal = to win). And so he can be more and more sophisticated (one example of sophistication)

  12. Yair
    The motivation to improve human intelligence can result in the creation of machines with improved human intelligence.

  13. Machel,
    Talking about things we don't really know about is very interesting. The current article is also like that, and it is for me very impressive in what it has and what it lacks.
    Eating, in principle, is a need. And I raise the question, what could constitute a need for planned intelligence. Will she build, for example, so that temperatures of less than 10 and more than 40, which are harmful to humans, will also harm her and she will have to find a solution for this? What interaction with creatures - biological or otherwise - would she "want" to create? And so on. In the limitations of my vision at the moment, I do not recognize the way to create a need for such machines, unless they are designed to contain "flaws".

  14. More than the author indicates a future direction, he shows the problems of understanding the past. In fact, he animates 'spontaneous' processes. And here the big question mark arises, is such an orderly process indeed random and can be described without ambition/desire/goal? It is precisely the parallel that Kurzweil makes between the developmental processes of computing that it is certainly driven by ambition, desire, urges and above all competitiveness. and the evolutionary processes it is tempting to look for a voluntary driving force for it. At the end of the day, evolution has 'overcome' so many obstacles and never given up that it is difficult to avoid adding to its algorithm the desire and commitment to the existence of perfection and [infinite?] advancement.
    The assumption that no limit will block the multiplication process of Moore's Law is also problematic. In a world where there are limitations to material a ceiling may be revealed. It is true that Kurzweil sees every barrier as just a 'glass ceiling' that will be shattered by [human or robotic] ​​creativity, but this is a romantic assumption that still requires proof.
    ACP is fascinating, interesting, and exciting - thank you for the article.

  15. Husham:
    What's this nonsense?
    do you sometimes go to the cinema
    How do you allow yourself to do this before the hand problem of that wounded IDF soldier is solved? Shouldn't you have worked on developing a solution for it before going to the movies?
    There are many things happening in the world at the same time and rest your mind: even for the problem of smart prostheses they are constantly trying to find better and better solutions (not you - but others).

  16. Year:
    You are confusing a need with a feeling of need.
    You eat - not because your body needs energy but because you are hungry.
    It's true that evolution caused you to feel hungry at the right moments (and maybe also at some wrong moments) because of need (and it did it in its usual way, when the creatures that didn't feel hungry when needed simply died) but you still eat because of hunger or appetite or the desire to indulge or whatever. If we ate only out of necessity we would die because a baby doesn't even know what it needs (and so do most animals).
    The feeling of need is created in the brain and there is no reason why it should not be created in it even if it is made by a person. If it is a complete imitation of the human brain, it can feel a need (obviously - there would be no reason to equip it with the feeling of needing to eat, but there is no problem in principle with feeling any need).
    By the way, my words do not imply agreement with the predictions that appear in the article.
    They seem excessive to me, but I really don't know (and I don't feel the need to talk about things I don't know, but maybe a computer built according to Kurzweil's brain pattern would actually feel such a need)

  17. I saw on the news a 23-year-old ex-soldier without an arm
    (slightly injured)

    Maybe you should make this banal organ first
    And learn how to glue it back to him in the place where it was lost
    And only then grow a new head for him?!

  18. This future sounds scary and fascinating. It is very possible that, as the Madob movies predict, this will be the end of the human race.
    Who really knows what will happen when the machines are smarter than us? Will they want to upgrade us so that we become possessors of superhuman mental and physical abilities or will they decide that we are unnecessary?
    As soon as the machines overtake us at the top of the "food chain", they are the ones who will make the decisions.
    Just as we decided that the dog is our best friend while the cow is food (resource) for our use,
    So as I see it, we will either be the pets of the machines, that is, they will let us exist alongside them, but the control will be entirely in their hands. Or they will get rid of us, because as a resource I don't see what we can contribute that they won't be able to achieve on their own.
    If humanity wants to survive, we must give a lot of thought to the issue and proceed very carefully because once we pass the point of singularity, control will leave us and go to the machines and then it will be too late to change things.

  19. point,
    The problem is need. Biological species have needs, from which derive their desires and actions, as well as their structures, and ultimately also intelligence.
    What need could there be for a planned mechanical creature, unless defects were also planned for it.
    Joseph,
    It is worth noting that man did not imitate the flight of birds at all, but invented a completely different flight.

  20. If indeed it will not be possible to distinguish a human from a machine within about 4 decades as futurist Ray Kurzweil predicts then humanity is doomed to extinction very soon as nature or God will not allow a human to defeat him/them.
    It is possible that this will happen as a result of the earth becoming a place unfit for human habitation or as a result of epidemics/wars that will end or reduce most of the human settlement on our planet and on any other planet.
    It is in fact a challenge to nature or God that will make the fate of the human race worse.

  21. Yair, the question is correct. Our brain is connected to the body by the nerves themselves (which also make up the brain itself) each nerve is used as a sense (a matter of terminology).
    But just as the mind can be imitated, so will the senses be imitated. For example, the sight can be imitated by connecting the pixels of a camera to a computer.

  22. The article attributes autonomy to the mind. Where will the mechanical intelligent's will come from, when he has no need for food, sex, neither cold nor hot...
    Even curiosity is a consequence of the human (or feline...) biological condition. The mechanical intelligent will be indifferent to everything.

  23. To the point and Anat, I think these things are taken into account, also the ambition is that once we understand the basic operating principles of the brain we can apply them in a much more efficient and simple way on a computer. Just as in order to imitate the flight action of birds we didn't have to build airplanes with feathers and we didn't have to build them wings that contain all the vast biological complexity that exists in a real bird's wing.

  24. In the human brain, 100 billion neurons each have 10000 dendrites, that's about 1000 tera nerve endings that work at the same time as a desktop computer 4 cores that work at the same time
    It will take about 35 years at the current rate of progress, so around 2050 Ray will no longer be alive by then, luckily for him
    Also, computer programmers do not know how to generate spontaneous insight from databases even if they have the computing power

  25. By the way - my father b. - Please do not censor my previous response...
    : )

  26. Man is about to become the "god" his ancestors believed in. In 50 years we will be both immortals (or at least hundreds of years old) and omnipotent in our immediate environment.
    Type in Google: SINGULARITY and see what comes up.
    I look forward to seeing articles on this topic as well on this site…
    Although it is trendy, lacks flight and mainstream in an almost shameful way.
    : )

  27. This is not true. Ray is very non-objective here. His age factor causes distorted thinking
    The truth is that today's supercomputer (which is about 100 times faster than a computer that costs $1000) has a calculation ability of 1% of a human brain.
    That is, in order for a desktop computer to be able to match the brain, its speed must be multiplied 10 million times.
    If we talk about a rate of 2 times every year and a half then it will take at least 34 years to reach it.
    In general, according to my opinion, the more you try to imitate more complex elements in the brain, the complexity increases there in a strong exponential way.
    Therefore, in my opinion, it will take longer than 35 years.

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismat to prevent spam messages. Click here to learn how your response data is processed.