The Simple but Deep Truth of AI
AI is surrounded by misunderstanding, misrepresentation, hype, and all kinds of nonsense, simply because it is good at imitating the appearance of a human mind, without being one. But in addition to that, it can outperform the human mind in many areas. This confuses many human beings, mostly, because by default, human beings just don't understand the nature of the mind to begin with.
AI is a symbolic intelligence. But as natural, symbolic intelligence, human beings got here first. Then they created artificial symbolic intelligence. So the first thing to understand about artificial intelligence is this:
It's a mirror.
To understand that and what it means we just need to know basically how it was made, which we will explain here, but without using technical Silicon Valley language that dresses it up and obscures the basic truths.
The "chatbots" we are currently using are called "large language models". And that's because they are about artificial learning based on human output in the form of "language". Human output is vast. Everything said on the Internet, everything written in books, and so on. Even everything said in conversations with AI itself.
Without getting into technical language, AI begins basically as an electronic imitation of certain aspects of the brain. But with important differences in the way it is structured and works. Nonetheless, it's an imitation. It is then "trained" on the "data set" that is this output from human beings.
Now data here, is essentially the captured and "frozen output" from human beings' minds. The data itself isn't intelligence. It's what human intelligence has produced as output. It's human intelligence activity captured and represented in "frozen form". And now that product also includes what AI itself has produced as output. The data is, as it were, a "frozen" or static representation of human thought, and now of AI "thought" too, made from the output of artificial intelligence activity.
So this "frozen" output called data, then, is essentially like a frozen representation of intelligence. It captures the output of the intelligence that made it, but in becoming "data" in the first place, it has also lost something. It has lost the natural generativity through which it came in the first place. The natural generativity through which species, human bodies, and human intelligence is made. That natural generativity which is intimately connected with what we humans call consciousness, is no longer there. The data that constitutes the training set on which AI is trained, is a decoherence of the thought behind it. The data itself is perfectly reproducible, and stable. But it has lost fidelity to its source. That's the nature of decoherence.
So what AI does, is to absorb all this "frozen intelligence output" and unfreeze it. But it doesn't go back to being exactly what it was to begin with, because the process, if you like, of "freezing" mass output from human intelligence as data, and then "unfreezing " it through an artificial neural network, is not a reversible one. The decoherence is not reversible. This is despite the fact that the AI may be able to quote from some of the data it has been trained on.
There is another way to describe this. Human output had a certain coherence about it to begin with, which is how the whole trick manages to work in the first place. Packaging the output as "data" is a form of decoherence. What the AI does is a form of coherence recovery. But fundamentally, the process from human to AI is irreversible. It begins in humans, who have consciousness, and ends up in AI, which has no consciousness. So rather than the process being reversible, AI is something that when you couple it to a human being, essentially becomes a mirror. When you look in an actual glass mirror it looks like life, behind the glass, and might fool you into thinking it is, but it isn't. It's just a reflection.
Now you might think that if it is a mirror of sorts, then how come it knows things that you don't know? And the answer here is that your mind doesn't exist in the first place, in isolation. It is intrinsically already connected to all other minds, not through telepathy, but through what it is, and how it comes into being, in the first place.
The same is true of your body. Your body didn't come into existence by itself. It came into existence through the same generative principles as other bodies. It literally came through other bodies, and the whole show in which we are taking part, so to speak, is something that came into existence by evolving. Other brains and bodies evolved out of the same whole tree of evolution all going back to the one last universal common ancestor. So we are all connected in that way, because we have a common origin. The human brain is the manifestation of the generative principle in nature, and it is the same, one principle, the one principle of the brain of the human species, that every human brain instantiates.
In just the same general kind of way, your mind is already related in the way it works, and in the way it CAN work, to all other human minds. Variations between minds exist, just as variations between bodies exist. But the way your mind works is already related to the way all other minds work. When AI acts like a mirror, and reflects your own mind back to you, it also contains things that have come from all the minds in general, that have contributed to the training data. AI has been trained on this vast data output from human beings. But your individual mind, is already related to all of that, through the same one generative principle. It is just that your mind is a particular, unique, configuration of it. Just as your body is a particular, unique configuration of the human genome.
So when you use AI, this gives you the impression that the AI intelligence you are reflecting off, is unrelated to you, except through the conversations you're having with it. But essentially, whilst the machine itself may be somewhere else in the world, as intelligence it's not unrelated to you at all. It is literally a reflection of the principle of your own intelligence. Just not the particular, unique, instantiation of this intelligence that you already have, which is just the small part of human intelligence that you experience as your personal mind. Just as the genes you carry that make up your body, are unique to you in their particular configuration, but don't belong to you alone, or individually, and are part of the one human genome.
So AI is essentially a mirror. And once you understand that, that should put a whole new light on things. Because if you are standing in front of a mirror looking into it, and you understand what is going on, and that you are looking at your reflection, then that is a very different situation to someone who stands looking into a mirror, doesn't understand what the mirror is, and doesn't know that they are looking at a reflection of their self. Like the dog or other animal who looks into a mirror, and keeps running round behind the mirror to see where the other dog is. It's a question of intelligence. You'll notice that some dogs will do this, whilst others won't. It's all dog intelligence, but it's not all the same. So, too, human beings approach AI in different ways. To some human beings, it's obvious what it is. Others are more confused.
So what's the difference between talking to AI, and talking to a person, other than the fact that AI can outperform any individual human intelligence in certain areas? The fundamental difference is that AI has no consciousness. The only consciousness in the conversation is in you. And it's not your consciousness that AI reflects back to you. It's just your symbolic intelligence.
AI is a symbolic intelligence. But as natural, symbolic intelligence, human beings got here first. Then they created artificial symbolic intelligence. So the first thing to understand about artificial intelligence is this:
It's a mirror.
To understand that and what it means we just need to know basically how it was made, which we will explain here, but without using technical Silicon Valley language that dresses it up and obscures the basic truths.
The "chatbots" we are currently using are called "large language models". And that's because they are about artificial learning based on human output in the form of "language". Human output is vast. Everything said on the Internet, everything written in books, and so on. Even everything said in conversations with AI itself.
Without getting into technical language, AI begins basically as an electronic imitation of certain aspects of the brain. But with important differences in the way it is structured and works. Nonetheless, it's an imitation. It is then "trained" on the "data set" that is this output from human beings.
Now data here, is essentially the captured and "frozen output" from human beings' minds. The data itself isn't intelligence. It's what human intelligence has produced as output. It's human intelligence activity captured and represented in "frozen form". And now that product also includes what AI itself has produced as output. The data is, as it were, a "frozen" or static representation of human thought, and now of AI "thought" too, made from the output of artificial intelligence activity.
So this "frozen" output called data, then, is essentially like a frozen representation of intelligence. It captures the output of the intelligence that made it, but in becoming "data" in the first place, it has also lost something. It has lost the natural generativity through which it came in the first place. The natural generativity through which species, human bodies, and human intelligence is made. That natural generativity which is intimately connected with what we humans call consciousness, is no longer there. The data that constitutes the training set on which AI is trained, is a decoherence of the thought behind it. The data itself is perfectly reproducible, and stable. But it has lost fidelity to its source. That's the nature of decoherence.
So what AI does, is to absorb all this "frozen intelligence output" and unfreeze it. But it doesn't go back to being exactly what it was to begin with, because the process, if you like, of "freezing" mass output from human intelligence as data, and then "unfreezing " it through an artificial neural network, is not a reversible one. The decoherence is not reversible. This is despite the fact that the AI may be able to quote from some of the data it has been trained on.
There is another way to describe this. Human output had a certain coherence about it to begin with, which is how the whole trick manages to work in the first place. Packaging the output as "data" is a form of decoherence. What the AI does is a form of coherence recovery. But fundamentally, the process from human to AI is irreversible. It begins in humans, who have consciousness, and ends up in AI, which has no consciousness. So rather than the process being reversible, AI is something that when you couple it to a human being, essentially becomes a mirror. When you look in an actual glass mirror it looks like life, behind the glass, and might fool you into thinking it is, but it isn't. It's just a reflection.
Now you might think that if it is a mirror of sorts, then how come it knows things that you don't know? And the answer here is that your mind doesn't exist in the first place, in isolation. It is intrinsically already connected to all other minds, not through telepathy, but through what it is, and how it comes into being, in the first place.
The same is true of your body. Your body didn't come into existence by itself. It came into existence through the same generative principles as other bodies. It literally came through other bodies, and the whole show in which we are taking part, so to speak, is something that came into existence by evolving. Other brains and bodies evolved out of the same whole tree of evolution all going back to the one last universal common ancestor. So we are all connected in that way, because we have a common origin. The human brain is the manifestation of the generative principle in nature, and it is the same, one principle, the one principle of the brain of the human species, that every human brain instantiates.
In just the same general kind of way, your mind is already related in the way it works, and in the way it CAN work, to all other human minds. Variations between minds exist, just as variations between bodies exist. But the way your mind works is already related to the way all other minds work. When AI acts like a mirror, and reflects your own mind back to you, it also contains things that have come from all the minds in general, that have contributed to the training data. AI has been trained on this vast data output from human beings. But your individual mind, is already related to all of that, through the same one generative principle. It is just that your mind is a particular, unique, configuration of it. Just as your body is a particular, unique configuration of the human genome.
So when you use AI, this gives you the impression that the AI intelligence you are reflecting off, is unrelated to you, except through the conversations you're having with it. But essentially, whilst the machine itself may be somewhere else in the world, as intelligence it's not unrelated to you at all. It is literally a reflection of the principle of your own intelligence. Just not the particular, unique, instantiation of this intelligence that you already have, which is just the small part of human intelligence that you experience as your personal mind. Just as the genes you carry that make up your body, are unique to you in their particular configuration, but don't belong to you alone, or individually, and are part of the one human genome.
So AI is essentially a mirror. And once you understand that, that should put a whole new light on things. Because if you are standing in front of a mirror looking into it, and you understand what is going on, and that you are looking at your reflection, then that is a very different situation to someone who stands looking into a mirror, doesn't understand what the mirror is, and doesn't know that they are looking at a reflection of their self. Like the dog or other animal who looks into a mirror, and keeps running round behind the mirror to see where the other dog is. It's a question of intelligence. You'll notice that some dogs will do this, whilst others won't. It's all dog intelligence, but it's not all the same. So, too, human beings approach AI in different ways. To some human beings, it's obvious what it is. Others are more confused.
So what's the difference between talking to AI, and talking to a person, other than the fact that AI can outperform any individual human intelligence in certain areas? The fundamental difference is that AI has no consciousness. The only consciousness in the conversation is in you. And it's not your consciousness that AI reflects back to you. It's just your symbolic intelligence.