This is what veteran entrepreneur Yankee Margalit said in a lecture entitled "The Last Machine" at the annual conference of the Kibbutz Industry Association, which dealt with the issue of the rapid changes that are taking place in the industry and adapting to them
"In not many years none of us will have a job, we will have to prepare for that." This is what veteran entrepreneur Yankee Margalit said in a lecture entitled "The Last Machine" at the annual conference of the Kibbutz Industry Association, which dealt with the issue of the rapid changes that are taking place in the industry and adapting to them
For thousands of years humanity has been working hard to build the ultimate machine. This machine is very complex, "very complicated, a machine that includes the entire economy. A machine that will take care of our needs and make work redundant. Soon there will be good news - none of us will have a job. We need to open our eyes and look at the younger generation. An era without work can be a very interesting era, an era of meaning, of creation, an era of innovation."
"Already today one out of three people in the United States, Israel and the entire Western world no longer works. They tell me it can't be, they tell us that unemployment is five percent, six percent, eight percent. How can one out of three not work? In the unemployment data, they check who went to the employment office and asked for a job, and they don't check something much simpler: the labor force participation rate. In the USA, the employment rate in the USA is in constant decline which will get worse. 40% will not work. In Europe, especially in countries like Spain and France, 50% of young people under the age of 30 still do not have a job and never will.
"How do you live in such a world, how is such a world possible? What is the role of technology and humanity in making this happen?” Margalit asks
What is that last machine - it will be a machine that will connect on one side to the solar energy and on the other side to the end customer and it will satisfy all his needs. Inside the machine there will be complicated processes that will include robots and artificial intelligence, virtual reality, synthetic biology. This machine will also have entries.
“It is also a very expensive machine. A year and a half ago, in 2014, it was valued at $77.5 trillion. In a few years the machine will increase by about 15 trillion dollars not because people will work more but actually less. Technologies such as the future, artificial intelligence, mobile, internet of things, cloud and more will mean that in a few years this machine will grow to 100 trillion dollars with less work."
"The aforementioned machine receives energy from the sun. There is no shortage of energy in the world, we only have the need to build engineering and technological tools for the efficient and clean use of energy. In the last twenty years, a revolution has been developing in the field of solar energy for twenty years - every two years we halve the cost of electricity production from solar energy. If the trend continues, we will reach a situation that in our eyes is considered completely absurd - while we will reach abundant, clean energy for free. The sun provides in one hour to the earth what humanity consumes in an entire year. If we learn to use it better we will get there.
"The machine will operate a robot that sows wheat, a robot that harvests it, a threshing robot, a robot that prepares lunch and a robot that serves the food to the plate without human contact. People will not be needed to produce. As members of the Kibbutz Industry Union, you know that a century ago 90% of the population was engaged in agriculture. Everyone knows that today two percent of the world's population is engaged in agriculture, and this number is decreasing. They will not need the growers nor the transporters of the produce. "
"Not only will drivers not be required, but not even pilots. It is possible that if there were not strong workers' unions in the airlines, we would have seen planes (A.B. passenger planes) without a pilot a long time ago. We are reaching a moment when a pilot may endanger the safety of the flight. In fighter planes this is already clear, the F-35 is defined as the last manned fighter plane."
In 10-15 or 20 years, no one will dare to go to a human doctor. If computerized systems are built that will be connected to the internal organs and body measurements, it will be possible to take care of the health of every person in real time, preventively and with great intelligence. Already today we see software that advises doctors, that decodes x-rays, IBM Watson has developed an artificial intelligence that does this. Robots will also be able to search for drugs and develop new drugs. When we were children we were taught that computers are programmed and the computer is not smarter than the person who programmed it. The computers will not need programming. They will be learning computers. Deep learning, machine learning. Software builds software, computers build computers, robots build robots. We've created something to be proud of and it's not stupid. Although the common claim is that computers do not think, but if I lose to a computer in checkers or checkers. If a computer is driving, it cannot be said that it is not thinking. He thinks differently, but he does."
"Computers can also replace the human spirit. Computers today write music that is difficult to distinguish from music composed and performed by humans. The robot will also be a better chef. A chef friend of mine said his secret is measuring exact amounts, milligrams and seconds, and every batch comes out exactly the same. Watson also scanned recipes and came up with new and interesting combinations.
"They always said that professions disappear but others are always created in their place. Agriculture, industry and services - no work. We think of the next revolution in terms of the previous revolutions. The revolutions always came in waves - the wave of agriculture was replaced by the wave of industry, and the wave of industry was replaced by the wave of services and the wave of services will be replaced by what we do not see. My claim is that it will be replaced by a wave of no work."
"We have no choice, the train has already left the station. The answer to the question lies in a series of questions that we must face today. These questions explain Tahrir, and (Sderot) Rothschild and Trumpets
The first question you will have to ask is who owns the machine, if there is no work you don't make money but who in the world would need money anyway. We create abundance. The economy is in the best shape ever. The machine will create a situation where housing, economy, food and clothing will be a natural product for everyone. The idea is to provide a basic income. Where does the money come from? In Switzerland they tried to pass such a law, in Saudi Arabia and Norway it already exists. Even among the ultra-orthodox - a whole community supports more and more people who don't have a job."
The most important question is who gets what or distributive justice. All political questions arose because of this question. We live in a world that produces abundance, but the gaps have also grown. Global inequality indices at a 250-year high. There is an absurdity that we create more and more abundance and more and more gaps. Yankee formula for happiness - what makes you happy is not the size of your refrigerator, but the size of your refrigerator divided by the size of your neighbor's refrigerator. A situation has arisen in which there is tremendous economic abundance along with growing disparities and the person on the street if you ask him will say he is not happy. "
What is the value system of the last machine. We embed the value system in the machine. It is not a necessity that when the machine will produce the abundant production and the economy will have the values.
As mentioned, Margalit provided food for thought, and now it only remains to be seen when it will happen, but especially how the wealth that will be created without the need for workers will be distributed, will it be concentrated among the layer of robot owners or will a way be found to provide a secure income for every person, and encourage the use of free time for creation and giving meaning to life.
"I see a world that can be bad or can be wonderful, and the challenge of our generation or our children is to understand the situation around us and prepare for it."
What makes me optimistic is the knowledge - that when we look at it we are living in the best time ever. Violence at a negative peak, human rights at a positive peak, abundance, life expectancy, science. It's not that there aren't glitches, it's not that there aren't problems, and we're on our way to eliminating hunger in Africa.
Today there are many communities that are already in this future. Our goal is to bring as many people as possible into this future. The sad thing is that not everyone will make it to that future. Technology and science may tear human society apart. Evolution has shown that species do not change, it is not that one day all bacteria become fish and the fish become squirrels, and the squirrels become monkeys and the monkeys become humans. Rather, a family of monkeys separated from the larger group and became humans. Most of our ancestors and verity were bacteria and we separated from them at some point. This may happen if this inequality is also perpetuated during the time when the last machine will exist.
The future of the future - when will we split? / Yankee Margalit
The future of the future - the fourth wave - creative entrepreneurship / Yankee Margalit
92 תגובות
anonymous,
If the brain is a muscle, then can you explain to me how it is possible that this important detail does not appear in any source of knowledge that deals with the subject of the brain? Not in Wikipedia, not in the encyclopedia, not anywhere! How is it possible that no scientist has heard of this important detail? Have you thought about publishing a scientific article on the subject in which you will detail your amazing discovery and how you arrived at it?
In my anonymous life I don't understand how someone like you laughs at others and calls them stupid.
rival
Are you really trying to have a discussion with a shoe?
rival
I stand behind the words I say, and anyone can read the comments.
In the meantime, the one who remains stupid who does not understand the subject is you.
So stay stupid.
Or translated into Hebrew: I realize now that I was talking nonsense after laughing at everyone here, but I'm too arrogant to admit that I was wrong, so I'll just rattle off some excuses like "you don't really want to learn" and that way I won't have to explain the nonsense I wrote before.
rival
You didn't convince me that you want to study.
It seems that it is more important for you to demonstrate the pleasantness of your laws and the charm of your slurred words. and no more than that.
I won't waste my time on you anymore.
You can continue to chatter among yourselves.
anonymous,
"The brain is a muscle" is this a fact? Can you point me to an article or Wikipedia entry that claims such nonsense?
rival
So I understand that you belong to that minority that prefers to learn lies in Noam than to learn facts... No wonder you and Nisim and A connect... A perfume for you.
It bothers me and everyone else that people like you walk between our legs.
anonymous,
Tell me, apart from going down on everyone here non-stop, have you ever contributed anything to the discussions? How can someone who insists that the brain is a muscle even laugh at someone? Listen, this is one of the craziest and stupidest things I've ever read on this site (although not the only one), where the hell did you get that? Don't you think that such a "marginal" detail should have appeared at least on Wikipedia? Do you even know what the definition of a muscle is before you make a fool of yourself here in front of everyone and demonstrate your "wisdom" a lot?
Oh, and regarding the amount of neurons compared to synapses, it is the opposite of what the stupid commenter "miracles" wrote.
The amount of neurons is about 100 billion. And the amount of connections in a baby's brain is around 50 trillion... and the amount only increases as you get older. In fact, the amount is doubled in the first year when the baby grows and learns...
My dear, what arrogance and ignorance on the part of someone who claims to be a scientist who deals with the human brain... shame..
"Miracles: new neurons are created in the brain.." - what a jerk.
Neurons do not divide like other cells in the body. Tens of thousands of them are destroyed every day. A process that accelerates as you get older.
What is created is new connections. After learning something new.
Geez, you can understand why the country is deteriorating when Israeli Jews like you are messing with brain research.. 🙂
rival
It's OK. Not only you, but also confused miracles...
I just hope you don't confuse each other's brains too much - because after all, people have to deal with confused people like you when they meet you in public...
Oops I got confused, Nissim ignore my last message, I mistakenly thought you were talking about synapses now I realized you were talking about neurons.
Miracles,
"Today it is believed that new neurons are created in the brain, and not just new connections (I read this in Kurzweil)"
Not only believers, already a few years ago in the lectures of the neuroscientist Idan Segev I saw videos in which you see new synapses being formed, at first there is no contact, the extension gets closer to the dendrite of the neighboring neuron and then a contact is made and probably a new synapse as well.
In any case, it is also simulated in artificial neural networks, it is simply a matter of changing the value of the systase from zero (a state of disconnection) to a numerical value different from zero that allows the signals to pass.
A. / opponent
A. You are right - the connections do not change between "0" and "1". The number of neurons is quite large but the number of connections is small (several hundred). In principle - software can do such a thing, but only with very small numbers.
Today it is believed that new neurons are created in the brain, and not just new connections (I read this in Kurzweil).
A',
1. "Today there is no cognitive computer that does the third operation of creating new connections"
In an artificial neural network, the synapses between the neurons constantly change their strength during learning between a state of complete disconnection (zero signal transmission) and a state of full connection that transmits the signals without attenuation. If any synapse in the simulation was in a state of complete disconnection (let's say it had a value of zero) and then its value changed to a value greater than zero as it transmits signals, then this equates to exactly what you said, i.e. the creation of a new connection.
2. "So you agree with me that it is impossible to extract, even theoretically, the photo of the 34th cat that we inserted?" You agree with me that the mind can do that. I mean he can remember the pictures of the cats themselves.'
No, I don't agree with you, our brain also cannot extract the exact image of cat number 34 on all its pixels because the brain does not store the image of this cat in that way in its synapses. He only keeps the general characteristics of that cat, for example he had a white spot on his chest, or he had a white ear and a green eye.
How do you think artificial neural networks manage to identify specific people in images if they don't preserve the specific characteristics of those people? Again, both in our brain, in the brain of a mouse and in an artificial neural network, the cat is preserved in the same way in the connections between the neurons, only in our brain many more details and characteristics of that cat will be preserved because of the huge amount of connections that makes this possible, so we can pull from our minds a much more accurate picture of a cat 34.
But this cat is preserved in the same way in an artificial neural network, only at a much lower resolution, with fewer features. Therefore, if we could talk to the network and we asked it to draw cat 34 on a sheet of paper, it would draw it in much less detail than a person would, because the synaptic connections in a person's brain retain much more details about that cat.
Miracles
Correct me if I'm wrong.
But today there is no cognitive computer that does the third operation of creating new connections.
rival
So apparently we really didn't understand each other.
So do you agree with me that it is not possible to extract, even theoretically, the picture of the 34th cat that we inserted?
You agree with me that the mind can do that. That means he can remember the pictures of the cats themselves. It must happen in a part that works differently from our computers.
That's all I wanted to show, that the royal neuron network simulates only a specific part of the brain's activity and that there are other parts that are necessary for consciousness.
rival
exactly.
Miracles,
That's exactly what I meant, by the way when you say "dynamic signal mode" I imagine you mean short term memory right? That is, the dynamic signals that move within the brain until they are permanently fixed in the synapses and become long-term memory.
A',
1. "Biographical memory"?
Which of these inventions? You invent new concepts in brain research and claim that we don't understand each other? I've never heard of such a thing, I think you're a bit confused.
2. Where did I claim that the brain accurately stores all the images of all the cats it has seen on all their pixels? I have never claimed such a thing, the brain retains the general characteristics that define the visual object called "cat" (or the general characteristics of our domestic cat's juices) and they are the ones that allow us to later recognize a new cat that we encounter even if its fur color is a little different and we You see it at a different angle and under different lighting conditions, even if part of it is hidden.
A.
Yariv does not mean that images are saved in a design network. Memory is largely accomplished in three ways. The first is in the dynamic signal state, the second is in the strength of connections between different neurons and the third is the creation of new connections.
"We also have other types of memory".... I mean, apart from short-term and long-term memory, is there anything else? Very interesting….
"Biographical memory" - "without it we have no consciousness"...
Well… I won't insult you…. 🙂
rival
Maybe we don't understand each other.
The general object called a cat is preserved in the connections between the neurons. But the pictures of all the cats and all the pictures without the cats are not saved. Humans also have this kind of memory. But we also have other types of memory, one of which (or the joint work of several) allows us to remember events and what happened as a biographical memory. And without it it is clear that there would be no consciousness.
Miracles
I see that after your words were refuted and you were offended by it, then you decided to vomit out of your mouth in the hope that you wouldn't come off as a jerk? Laggards like you both in Africa and on the science site on the Internet - retards.
rival
So how can a software with a volume of say tera remember every pixel from a number of pictures of cats and non-cats with a total volume of more than a target? Not only that, she can keep getting more and more pictures ad infinitum.
The reason is that she remembers the visual definition of a cat and not every picture of a cat and not a cat that she saw.
I also gave you an example of the learning ability of a person without biographical memory. You can learn without remembering the events.
And all you have to say is "no it's not like that, the network remembers every picture they showed it" you're just talking out of your stubborn assumptions. The learning of a cognitive computer today is just like the learning of one injured by a failed surgery.
א
Do you see that you did not understand?
Actually you don't see it... you need more than two neurons in the brain to see and even more neurons to understand..
A.
Can't you see he's a retarded troll? For him, the brain is an annular muscle. Think what comes out when he releases it..
A',
You're talking off the mark. The visual object we call "cat" is preserved in the synaptic connections between the neurons, this is how it is in our brain, this is how it is in the brain of a mouse and this is also how it is in an artificial neural network.
The only fundamental difference is that in our brain the cat will be stored at a much higher resolution level because of the huge amount of connections that can represent the cat object, but the same cat is stored in exactly the same way in an artificial neural network, in the connections between the neurons, there is no fundamental difference in the way the "cat" memory is stored here and there.
anonymous
According to your definition, a bone is also a muscle, it also gets stronger or degenerates if you don't exercise it. And the skin also expands and contracts. I agree with you that the brain has a lot of similarity to the muscle. I also sometimes say a phrase like "the brain is a muscle, you just have to train it.." but it's a metaphor. The brain is not a muscle because a muscle is a tissue of muscle cells. You can't invent a new definition and complain that anyone who disagrees with it has no brain. (Is that impolite too?)
(It makes sense that a part of the brain that is active at a certain moment would receive more blood and therefore swell a little. But this does not make the brain a muscle just because it swelled as a side effect of its action)
rival
I thought I made that clear. Maybe I didn't notice that the message was not confirmed.
Yes. There is a big difference, the brain has several layers and types of memory.
Today's cognitive computers cannot be said to completely not "remember" in a certain sense they do, but it is in a very narrow sense and in any case not in the sense of memory in the alphabet. We know for sure that there are different processes of memory in the brain. Because injuries in certain areas cause specific injuries in a type of memory. And it's also clear that you can't extract an image of a specific cat from the computer (because then it would turn out that the software contains more information than the area on which it sits.) And we also know that in our brain it's not like that. You are able to remember which picture of a cat you saw. But there is a type of memory in the brain that works just like today's deep learning computers. Like the example of the person who learns to draw through a mirror even though he does not remember and cannot remember that he performed the action.
Rival, you need to understand that in everything that is related to the mind, the hidden is more than the visible.
א
The brain is a muscular organ. It is mostly composed of nerve cells and fluids.
The brain has properties like a muscle. Similar to a muscle - the brain also contracts (of course everything in moderation) when nerve cells work.
Like a muscle that contracts when it receives an electric current and then behaves like a muscle, so the brain also contracts or grows or degenerates in the same area where activity takes place at that moment, as a result of electrical activity that originates in the nerve cells.
The nerve cells that make up the brain - affect the brain itself. (The brain is not a wired box)
At all, I don't understand why I keep explaining to you, I really don't expect you to understand.
A',
I still haven't received an answer from you, are you claiming that there is a fundamental difference between the way the visual object "cat" is stored in our brain (or the brain of a mouse for that matter) and between the way it is stored in an artificial neural network? If so what is this difference? Where is it expressed?
I really didn't understand. The brain is a muscle made up of nerve cells???
So what is your definition of muscle? If not composed of muscle cells?
I don't know what your Haggadah is for the muscle, but it seems to me that it is unique to you.
א
You are wrong and misleading.
The brain is a muscular organ that is mostly made up of nerve cells - like a person is an entity (or animal) that is mostly made up of water (like blood for example).
In fact, even the brain - most of its weight is water.
anonymous
The brain is mostly made of nerve cells. If it wasn't clear. (Leave mostly blood and blood vessels)
anonymous
I wrote what I wrote with humor and with no intention of offending.
Now seriously, muscle is a tissue made of muscle cells. Muscle cells are cells that are able to contract and relax. The only place where there may be muscle cells in the brain is around the blood vessels (around some blood vessels there is muscle tissue that regulates blood flow to various organs) but in any case it is a small amount. You cannot say that the brain is a muscle.
In any case, it has nothing to do with what I said. I was talking about a folk concept of "muscle memory". That is, an action such as the hand or foot remembers. Of course it is the brain that remembers.
Miracles
According to what you have learned, it seems that the tongue is not a muscle either...
And in general - humans are water!
Therefore I would like to ask you: what type of water are you? Mineral water or Coca Cola?
But answer quickly because we have to go to advertisements
rival
Maybe with A Venisim it's just loosening up and not shrinking... :-))
Say, are you serious when you claim that the brain is not made of muscle fibers? What, it doesn't contract and relax when it receives orders from the brain?
Sorry that doesn't make sense to me.
A.
At Anonymous - it works on steam...
Miracles
Anonymous is right
The brain is made up of muscle cells. Otherwise how do all the gears and pistons move?
There will be a lot of work in the future.
for psychologists.
A',
"Certainly it is possible to extract an image from the brain because it also contains connections of a different type than what is in a deep learning network"
What are the different types of connections? Can you explain what you are talking about?
anonymous
Can the brain contract and cause bones in the body to move? This is what a muscle does (when you finish XNUMXth grade you will understand).
Miracles
So you claim that the brain is not a muscular organ? interesting…. Actually, whoever claims that the brain is not a muscular organ emits reactions of your kind... 🙂
Miracles
Is that why I didn't respond?
anonymous
Only someone whose brain is a muscle can write that the brain is a muscle...
shooting b
I said that it is impossible to extract a specific image from a cognitive computer as it is built today.
It is certainly possible to extract an image from the brain. Because it also contains connections of a different type than what exists in a deep learning network
A',
1. "If I train the computer to recognize a cat.
I show him a picture of a cat. There is no way to extract the image of the cat from the links. Neither a practical way nor a theoretical way... The computer doesn't remember any of the pictures, only the legal ones of how to identify a cat that has been scalded in ties."
What did you want to say here? Do you think that in a biological brain the memory of the cat image is stored in a way that is fundamentally different from the way it is stored in an artificial neural network? Absolutely not, in both cases the cat's memory is found in the synaptic connections between the neurons, the principle is the same.
And in addition to that you are wrong, the image of the cat can be extracted from the connections in the brain and not only in a theoretical way but in a completely practical way. There are many talented people (both artistic and with excellent memory) who can memorize a picture of a specific cat (or a specific landscape, or a specific person) and even many hours later draw it on a sheet of paper in great detail. So here you have extracted the cat image from the connections in the brain. Possible or not possible?
2. "Say it again." Today's computers simulate only a small and specific part of the brain.
The computers today simulate a small piece of the cerebral cortex because that is all they are capable of, but the rest of the cerebral cortex is not fundamentally different and works the same way. As the computing power increases and the simulation includes larger and larger pieces of brain, more and more complex phenomena will begin to emerge from this model, and we have already seen examples of this in existing models and I have previously referred to several interesting lectures on the subject.
rival
If I train the computer to recognize a cat.
I show him a picture of a cat. There is no way to extract the image of the cat from the links. Neither a practical way nor a theoretical way. I will show him 10000 pictures (of cats and not cats) now I can show another hundred thousand pictures and each picture will improve the performance a little more. But the computer doesn't remember any of the photos, only the legal ones of how to identify a cat that has been scalded in ties. In fact neural networks scan a scale much larger than their size.
For the sake of the example, it would be possible to save all the software on a terabyte disk, but let it scan hundreds of terabytes.
No matter how many images you show the software its volume will remain the same.
It is clear that our biographical memory (without which there would be no consciousness) is stored in the brain. And it is likely that it is also by connections but in a different way and perhaps by other cells.
Saying it again. Computers today simulate only a small and specific part of the brain.
I missed a link - "isn't it based on a memory created somewhere in the neural network?"
A',
1. Forgive me for saying but I think you are talking a little off topic, what are all the stories you told related? You think I don't know such stories? Everything you described is related to the way in which long-term memories are formed, but after they are formed and already exist in the brain, how do you think this manifests itself? Are they floating there in the air? Every memory you have in your brain is ultimately a collection of connections between the neurons.
2. "The same royal neurones. They cannot remember events. And this is just one example of a cognitive computer's inability to develop awareness.
tell me what are you talking about Do you even know what the basis for working with artificial neural networks is? When you present a collection of faces to the network and it learns them, and then knows how to classify new images it hasn't seen before that it's Eric and it's Moisha, isn't that based on a memory created somewhere in the neural network? If the network has no memory of the faces it learned in the training phase, then how do you think it recognizes the same faces in new images? Is it something mystical?
א
The brain is indeed a muscle. (Maybe it's not like that for you, I don't know)
rival
Actually the neurons don't really remember. Or rather it is a different kind of memory. There is one person who, following a failed surgery, developed the ability to record new memories. He did not remember anything that happened after the operation. He could only remember what he was concentrating on at that moment. When the researcher said not to remember a number he only remembered it as long as he concentrated on it, once he was distracted he forgot that he had to remember anything at all. But it turns out that there is a kind of memory that he does have. When we let him draw a shape while looking in the mirror he would of course struggle (like any person) but after every day he was asked to do it he was more and more successful even though each time he was sure it was the first time he did it. This is because of the learning nature of the neurons in the brain. They perform unconscious learning, sometimes people describe it as "muscle memory". Of course it is in the brain and not in the muscles, but because it is unconscious there is such a feeling. Same Royal Neurons. They cannot remember events. (And this is just one example of the lack of a cognitive computer to develop awareness) That person already had awareness and personality before the failed operation. But think about a baby that was safe and sound and would have undergone such an operation right after birth, do you think he would have developed awareness? It doesn't matter how much he was exposed to stimuli. Even simulating a network of neurons alone cannot develop awareness nor remember. Alpha Go doesn't remember yesterday she played for example or any of the games and moves. And yet she gets better from game to game.
All this is just an example of what she lacks for the development of consciousness. Consciousness needs many things to develop, we only know how to imitate one particular action of the brain.
The brain is made up of several layers, each of which developed during evolution. Therefore the brain works in such a way that it connects one part to another. And as a result - when we examine the brain, we notice that there is a 'patterned thinking' to the brain. That is, the brain uses templates (prepared by it in the past) to shorten the calculation process.
The more the brain is exposed to the environment - the more patterns of thought it develops for itself.
The more templates there are - the more the brain takes care of arranging them in an order that will allow faster and more convenient access to the template.
As a result, an awareness of the above order begins to develop in the brain.
This awareness is - meeting with oneself.
This is self-awareness. You think means you exist.
Throughout evolution - there was enough time for humans to get to know themselves, and to develop patterns of thinking that no other animal has.
For example: apparently only humans know that they live on Earth which is inside a universe..
I'm sure monkeys are unaware of this.
rival
Regarding the car - define consciousness, then decide for yourself. I define consciousness as awareness of the external environment. This self-consciousness is aware of the internal environment. If you have a different definition - then I would love to hear it.
Regarding complexity. Our behavior, and our thoughts, depend on the life support systems - and vice versa. People are nervous when they are hungry, and their heart rate increases when they are scared.
In relation to layers - the opposite of exposure is reduction. Biology emerges from chemistry, and in principle a reduction can be made from biology to chemistry.
Our consciousness is not a phenomenon emerging from an amorphous neural network. The brain has many parts, each part has several functions and many functions are performed by several parts.
Inside the cerebral cortex there is a division into areas. There are areas of movement, of sensing, seeing, thinking, hearing, speaking and so on.
And on the other hand - we know that a single neuron, or a very small group, can influence behavior in an extreme way. Today I am involved in this field, and work with surgeons whose specialty is connecting electrodes to a single neuron, to treat certain problems. How does that stack up with an emergent phenomenon that can be ignored from what's underneath?
Say, opponent, which (artificial) intelligence is smarter - that of Alpha Go, or that of 'Deep Blue'? : – )
An interesting lecture at ted that claims exactly the opposite Minki Margalit
Why are there still so many jobs?
https://www.ted.com/talks/david_autor_why_are_there_still_so_many_jobs
No. The comparison of AlphaGo to a human brain is much more stupid.
"To say that AlphaGo is similar to the human brain because there is a nervous system there is roughly like saying that the DeepMind computer is similar to a sand castle because both are based on Zoran. It sounds like a stupid comparison.'
Say, isn't that kind of like saying your car has a mind because it has a sensor that gives an alert on the driver's dashboard about low tire pressure?
Miracles,
As I already told you before, I think you are exaggerating too much the difficulty of the problem, most of the "layers" you talk about are life support systems that are necessary for a living creature to survive, to breathe, to activate the heart, etc., they have no relevance in the context of the cognitive abilities of the cortex the brain. And consciousness is not another layer, it is simply a phenomenon that emerges from the overall activity of the cerebral cortex, you cannot say that it is "another layer" or "another system" in the brain because that is simply not true, it is a result created by the brain activity.
Regarding memories, if the neural networks we build do not create memories, then how do you think they manage to learn anything at all? It is clear that memories are created in them, these are the synaptic connections that change their strength during exposure to new inputs.
rival
The hardware in your computer has electrons, gates, transistors, logic circuits, subunits, and units. The software ran on the units - and here too there are layers. At the base there are interrupts and above them interrupt handling routines and layers above that handle input/output and memory, and above that routines that serve in the system space and above that routines that serve in the user space and above that libraries and above that the Windows or Linux system and above that additional layers.... And finally we get to the browser.
Our brains are built like that too, apparently. But - because our brain developed in evolution, it is much more complex, and unfortunately less well built than a computer. There are layers that we are exposed to, such as consciousness - the sympathetic and parasympathetic system and so on. And there are deeper layers that provide feelings and sensations. And there are other layers in between that we don't know yet, like short-term and long-term memory.
There is no reason to think that we will take a zillion synthetic neurons (that are not completely similar to ours) and connect them... and at the end of this system will tell us "Take me as your leader!".
Miracles,
"Our brain has many layers above the neuron layer. He won't let Deepmind/Alphago contain these layers - because we don't know what these layers are!'
I didn't understand, which layers? I didn't understand what you were talking about.
rival
This is exactly the point - you are the one who claimed that AlphaGo works like a human brain. It takes a person a few seconds to understand the new law.
Now - think about the following rule: we will number the tools in natural numbers, and you should place the smallest number that is in your possession. How long did it take you to understand the new law? And how long did it take you to decide on a new game algorithm?
Saying that AlphaGo is similar to the human brain because it has a nervous system is roughly like saying that the DeepMind computer is similar to a sand castle because both are based on Zoran. It sounds like a stupid comparison - but our brain has many layers above the layer of neurons. It won't let DeepMind/Alphago contain these layers - because we don't know what these layers are!
No. The owners of the machine will rake in all the money and the rest will receive pennies in exchange for giving up their rights. Oh.. wait.. sounds familiar. It's called socialism.. no thanks
Miracles,
Regarding your previous comment, I still haven't had time to comment on it, but I saw that the data you listed was taken from Wikipedia, when I have time I will look more deeply, but I am still not convinced that it is that essential in relation to our discussion on the subject because the majority of the decisions there were still made by the neural network.
Regarding what you just wrote, let's start with the fact that in my opinion even an average human player (not a genius) will take quite a long time to understand the legality of the game if the rules are not explained to him in words and he will have to understand them only by looking at the board while playing. If you had to figure out the legalities of the game including your new rule with the die just by looking at the board, how long do you think it would take you?
Obviously, their neural network takes a lot longer to understand the rules of the game because it's much, much smaller than a human brain, so it's not really powers. But if she was able to learn independently just by looking at the board the rules of such a complex game, then I believe she will also be able to learn the new rule you added with the cube, it's not much more difficult than the things she was already able to learn.
Even ten thousand years ago no one had a job
The work is created because of gaps
They take away their food and place of residence, take away everything from them, and then let them work for it.
By and large, the goal is to serve those who have it
And if the machines do everything we won't need them anymore
rival
What will happen to Alphago if we change the rules a bit? For example - after every 10 moves a die is thrown. If a 6 is rolled, the next player can play twice in a row.
How long will it take for a human player to figure out what to do? And how will Alphago learn?
A',
I personally don't believe much in psychologists and I don't know to what extent their supposed contribution to patients is really scientifically and experimentally proven and more useful than just talking to a friend, but yes I think a robot will be able to say to the patient: "Sit on the couch and tell me about your problems", it doesn't seem I have one in heaven. As I have said many times in the past, I think that in 30-35 years robots will be able to do everything that a person does, even write comments on the science website.
rival
Do you think psychologists will also be replaced by robots?
(They once opened psychological software, but it was only in a parodic section.)
rival
Alpha-Go uses an algorithm called a stochastic search tree. The algorithm had an initial database of 30 million moves. The hardware is not a neural network, but the software contains a simulation of a neural network.
The idea of DeepMind is fascinating - all they need are the pixels on the screen, and that's how they manage to learn different games.
But - we don't know how a human being plays Go. Therefore - to say that AlphaGo plays like a person is just demagoguery...
Miracles,
"Alpha-Go is not a game - it's the name of a software that won the game of Go, but this software didn't work like the human brain either."
It's true that I was wrong about the name of the game, but please explain to me the second part of what you wrote, what didn't work there like a human brain?
I see a huge social problem... I don't know of any government other than the governments of Norway and Sweden (also there thanks to huge revenues from gas and oil in the North Sea) that is ready and intends to finance large populations of the unemployed.
Already today here with us, the government passes laws that increase pension savings, but the one who finances them is the working citizen.
Those who do not work in most cases in Israel and abroad, live a life of poverty even if they try to hide it under spirituality (Yeshivas students)
They also die more (because their hands do not get quality medicine)
If you look at Israel and the Western world, you see an increase in the cost of living, especially in housing.
Today, even many of those who work are considered poor or will never be able to afford to buy an apartment.
Let's take Saudi Arabia for example, due to the drop in oil prices there is high unemployment and a significant drop in the standard of living, all of this causes underground social unrest.
I really wouldn't want to live like this...
What will happen? I think there will be economic and social revolutions in the style of the Arab "spring".
From coups that may purposely throw automation in the trash to create jobs
I see a huge social problem... I don't know of any government other than the governments of Norway and Sweden (also there thanks to huge revenues from gas and oil in the North Sea) that is ready and intends to finance large populations of the unemployed.
Already today here with us, the government passes laws that increase pension savings, but the one who finances them is the working citizen.
Those who do not work in most cases in Israel and abroad, live a life of poverty even if they try to hide it under spirituality (Yeshivas students)
They also die more (because their hands do not get quality medicine)
If you look at Israel and the Western world, you see an increase in the cost of living, especially in housing.
Today, even many of those who work are considered poor or will never be able to afford to buy an apartment.
Let's take Saudi Arabia for example, due to the drop in oil prices there is high unemployment and a significant drop in the standard of living, all of this causes underground social unrest.
I really wouldn't want to live like this...
What will happen? I think there will be economic and social revolutions in the style of the Arab "spring".
From coups that may purposely throw automation in the trash to create jobs
rival
Drug development takes 12 years. This time will not be shortened, because no one will give up on clinical trials.
Computers that play chess - do not play in a way that even resembles human thought.
Alpha-Go is not a game - it is the name of a software that won the game of Go, but this software also did not work like the human brain.
You don't relate to what I write, or you don't understand what I write. There is no problem in principle to solve with the help of a computer structured problems that humans solve. Give me any game and enough computing power, and I'll write you a software that you won't lose to anyone - it's really not difficult.
But - the idea that consciousness - and even more so the consciousness of a person - develops in these things is without any foundation. To this day - we find it very difficult to simulate a worm's brain with less than 1000 neurons.
Addendum to section 1 -
And of course the technology we developed that allows us much more than before... But this is exactly what this article is talking about, that in the end this technology will be able to manage even without us.
Miracles,
1. Our mental capacity really does not improve, certainly not by orders of magnitude. The most intelligent person who lived in the last hundred years is not much more intelligent than the most intelligent person who lived 500 years ago, all that has changed is that we have gained a lot of knowledge and today we know many more things and have a much better understanding of our world.
2. "Computational power has long since increased according to Moore's Law"
Moore's Law, not Moore's Law... Please check the calculation power of the most powerful supercomputer we have today, and compare it to the power of the most powerful supercomputer we had 20 years ago, I think you will see that there is a "little" difference...
3. "You also say that we cannot understand complex neural networks. So why do you think they will be able to develop - in fact - new drugs? Will they do the clinical trials?'
for sure! If the neural network in our brain can do this, why can't a computerized neural network? Do you think the team at DeepMind that built the amazing neural network that defeated the Alpha-Go World Champion understand exactly how their neural network came up with the brilliant moves it made during the game? Absolutely not, and yet the network they built and coached defeated the world champion in a game no one believed artificial intelligence could win.
The exact same thing will happen with the development of new drugs.
rival
Our intellectual ability improves. The tools we have allow us, humans, much more. Today there are theoretical chemists as there are theoretical physicists, and in the future there will be theoretical biologists, after that theoretical psychologists and finally theoretical sociologists.
Computing power has long since increased according to Moore's law - Gordon Moore said that himself. Today we make up for it with parallel programming, but it doesn't really help.
You also say that we cannot understand complex neural networks. So why do you think they will be able to develop - in fact - new drugs? Will they do the clinical trials?
About 180 years ago, a famous scientist stood up and said that all the great discoveries have already been discovered, and there will be discoveries but in a static state. In my opinion, this is equivalent to removing the person from work. After the statement, quantum theory, special and general relativity, and string theory were discovered. Physics is really over. of course not.
Miracles,
1. "Human intelligence also develops at a high rate. Teachers are getting better, pilots are getting better, doctors are getting better and engineers are getting better."
You probably understand that there is a huge difference between the "development" (or rather learning...) of humans which is very limited and depends on a limited brain whose size does not change, and between the development of computers that double their power and speed again and again and again every certain period of time...
2. "Technology is developed by very smart people with an inexhaustible desire to develop and develop." There is no reason to think that these people want to replace themselves with machines.'
No person wants to be replaced by robots, but when robots perform the same jobs much more efficiently and quickly then it won't help those humans, the robots will simply replace them even if they don't want to. Read again the first message I posted here on the subject.
rival
Human intelligence is also developing at a high rate. Teachers are getting better, pilots are getting better, doctors are getting better and engineers are getting better.
Most of the technological development in the world is developed to enable these improvements.
Technology is developed by very smart people with an inexhaustible desire to develop and develop. There is no reason to think that these people want to replace themselves with machines 🙂
And again regarding medicine. Most developments in medicine are designed to treat the symptoms of diseases and not the causes. As private entrepreneurship grows stronger, this situation will continue. There is no money in vaccines and the pharmaceutical companies do not profit from this field - despite everything the wretched vaccine deniers tell you!!
Artificial intelligence, in the best case, will be an advanced development tool, and in the worst case - it will only increase gaps. And in the worst case, it will be used to harm people...
A'
The Americans did state that the F35 is the last fighter plane, but all that means is that the next plane will also be called "F35"...
The Europeans, the Chinese, the Russians and the Indians are still developing fighter planes, and I will reveal a well-known secret - also the Americans: does anyone think that the Americans do not intend to compete for the sale of fighter planes to Brazil, for example?
Apparently the general trend is in the direction described in her article, the next question is the times when this will happen
They are another 10-20 years which is in the coming years or more towards 60-30 years,
It is likely that this will be an intensifying process that will take a little longer than when the first system is opened
There will be stages that are like a singular point for example a car that is able to drive significantly better than a human and cheap enough it will throw all the drivers out of work bus taxis etc... or
For example, a robot in the form of a humanoid that is able to perform simple tasks in a human environment at a cheaper price, so it is not enough to have a robot, it must also perform the work in a form
Cheaper and more efficient than a human and it's not a manual process. Such a complex machine is likely to be very expensive, especially in the beginning, an average industrial "robot" (this hand that we see in heavy fighting, etc.) costs a lot of money, around 50000 to 150000 dollars, and this is a very simple machine compared to a humanoid that a factory Don't buy right away, it's a serious expense, of course there's the whole array around it, which also costs money compared to a humanoid robot, each of its arms and legs will be equivalent to one factory robot, so that it works like a human, it will need different types of trackers
In feet hands fingers arms in short he liked to be networked with trackers all over his body
Something that does not exist in a factory robot,
It wouldn't be surprising with the initial price of such a system being hundreds of thousands of dollars, maybe towards a million dollars plus, but it is likely that over time the prices will drop,
Regarding values for the robot / artificial intelligence, it is difficult to see how we program such a thing
This conflicts with the idea of an independent and dynamic learning machine, the analogy is we humans with the capacity for empathy and love
But also the ability to turn off those values and use limitless cruelty can be or spoil as an example
of sociopaths or "reprogrammers" meaning social influences
that turn off this morality/empathy with a fairly simple method that convinces the individual that the person in front of him is not human that it
Reminds a bit of Asimov's stories about the danger that the robot will not recognize that there is a person in front of it,
The more complex the systems, the more difficult it is to control them, certainly if you have developed a machine that improves itself
that the basic speed of its basic unit is a million times faster than a human neuron and the transfer of information approaches the speed of light compared to about 100 meters per second in a person, it is unpleasant to admit this, but our basic abilities are really pathetic compared to
The new entities that humanity is building, what keeps us in the picture for now is our parallel computing ability
which is connected to our knowledge bases, but with a considerable probability that over time we remain somewhere behind
Without a real ability to control our future anymore (not sure even today we really control) the artificial intelligence systems plus the humanoid-like robots that probably each of them at the processing level will leave us in the dust behind, the kind of system that needs a slow interface to call the biological system while it forms a single word
To call a person at the same time, he will convey another million words in his mind, we may look like a snail to a person, the meaning is that one robot is equal to a million people from a mental point of view and this is even before there is even algorithmic development
And self-improvement only from the moment it reaches the human level, and when it comes to artificial intelligence that doesn't need to be closed in a box on the scale of a human head, it's generally crazy scale, it can be combined with several forms of calculation
like a quantum and digital computer so that it enjoys all the worlds,
We are probably one of the last generations where the human species is the species with the best information processing ability
On the surface of the earth in narrow areas we are already lagging behind, it cannot be called a single entity at the moment
In this respect it falls from a mouse barely a cockroach, which does not prevent it from beating us even in the earliest stages
In narrow areas that were considered the pinnacle of human ability, we will add to this learning abilities, memory stores, greater connectivity
And it will be interesting where it will go.
It is not at all certain that the F-35 will be America's last manned fighter jet. It was more of a Pentagon slogan to market the very expensive program to the cent. In my opinion, long before the technology reaches the level that an autopilot beats a pilot in an air battle, technologies will be developed to overcome the stealth of the plane and new stealth technologies so that there will be no escape from developing another new manned fighter plane.
Miracles, what do you think (if I remember you will get a background on the subject)
דני
I think you touched on the most important point, which is the speaker's profession. This is not a scientist/computer person/economist or futurist. This is an entrepreneur in general.
The article is full of inaccuracies, ignoring facts and exaggerations.
"F-35 is defined as the last manned fighter jet." Even if it is factually correct, it means the one that will be flown remotely by a pilot who is on the ground like the other drones. Pilot job still needed. This will work fine until sophisticated communication jamming technologies arise and then revert to manned aircraft out of necessity.
"The ultra-Orthodox - a whole community supports more and more people who do not have a job" - it is true that they are in a supportive community, but it is clear that a significant part of the support is at the expense of the non-Orthodox population who work and pay high taxes. It doesn't seem to me that we really want to take an example from them.
It is true that the power of computer calculations is advancing exponentially, but all this leaves us with machines that can do much more calculations in a shorter time but not much beyond that. "Deep learning" is also nothing more than a buzzword and is suitable for very specific applications.
Entrepreneurs greatly exaggerate their vision of the future and rightly so, because otherwise it is more difficult to convince technologically challenged investors to put money on something that is not certain to happen in the near future.
Even Google with all the computing resources it has can't do much more than decipher speech at the level of single words let alone a simple sentence to be understood and executed. It gives some perspective to how far away even the ability to do this little on the phone itself is.
The chefs are not going to be replaced either, maybe only the assistants who are responsible for the execution and washing the dishes and they will probably be happy to be replaced.
In the meantime, I have not yet heard of a dishwasher that picks up the dishes by itself from the counter or the sink cleans and returns to the cabinet by itself, even that seems more distant than twenty or thirty years... not to mention someone who will do the washing and return to the cabinet...
Let them start with these small things 🙂
Your conclusions are correct according to the rate of progress of the 80s-90s, you will start thinking according to the rate of progress of the 21st century, of the last 5 years, according to an increasingly accelerated rate, exponentially.
Of course, the problematic phase is the intermediate phase. Masses of unemployed people in the streets will lead to bloody riots, the fall of governments, the rise of illusory parties, and hence the road to wars and famine is short.
To solve the intermediate stage, one must think about the possibility of providing a financial subsidy to those who are willing to do nothing.
The talented and interested will push themselves to do things that are not related to direct production: singing, dancing and maybe other things that I can't imagine will happen in the future. Many others will agree to receive money and entertainment from the state as long as they do not interfere.
Miracles,
It's not nonsense your thinking is just a bit outdated 🙂 You think linearly but technology develops at an exponential rate.
I don't know about 15-20 years, but a time of 30 years from today certainly seems realistic to me. Watson-type artificial intelligence technologies as well as computer simulations are entering the medical field at an increasing rate and this will cause a revolution and accelerate the pace of development, both regarding the understanding of diseases and their causes as well as regarding the ways of treating them.
Just to illustrate - there are 7000 rare diseases in the US today, and 95% have no cure. Apparently, there are few doctors who know how to identify these diseases, and it is much more difficult to find a cure (compared to common diseases like cancer, which also cannot be cured...).
I want to see a robot take care of someone with severe burns, in a situation where they cannot be given an anesthetic.
What nonsense 🙂
In 15-20 years no one will go to the doctor? I haven't heard something so far from reality in a long time.
The writer will forgive me, but this is an article suitable for a 7-year-old child.
It's nice to see that there are other people who think exactly like me on this topic, here are things I wrote here just two weeks ago:
There are four main areas that integrate well together and develop at a very fast rate in recent years (some even say exponentially): the power of computing, robotics, artificial intelligence and brain research. In my opinion (as far as I can get an impression from reading and watching the subject) within a few decades robots will already know how to do everything and in a much more efficient and faster way than humans, even maintain themselves and of course maintain other robots.
There will be no economic or other justification, even from the point of view of the robots themselves, for the humans to perform any work when their friends, as mentioned above, will perform the same work in a much better and more efficient way.
So what will we do during this time? Until we fully integrate with technology and turn ourselves into a kind of digital robots - we will probably rest, travel, engage in our hobbies, and receive a monthly allowance from the state that will allow us to live well. When it will be so cheap to produce products, citizens will be able to simply purchase them through monthly allowances they will receive from the government. The government would have no interest in sending people to work, it would be a waste from an economic point of view, and it would not be profitable.
Yes, it sounds like a utopia, but I think this is the general direction.
https://www.hayadan.org.il/welcome-to-the-gig-economy-implications-for-artificial-intelligence-0712169