Artificial Intelligence – Lesson 3 – Rabbi Michael Avraham

[Rabbi Michael Abraham] Good. We’re in the preliminaries leading up to the discussion of artificial intelligence. I tried to define a bit some different characteristics within this creature or structure that we call a human being, and at the beginning I spoke about the question of vitalism. The question of vitalism basically tries to check whether biological matter is a different kind of matter from inanimate matter. Today, among biologists, it’s accepted to think not; this whole discussion even strikes them as a bit laughable. The second question is the question of dualism, which is often tied to the question of vitalism, but in truth it isn’t connected. The question of dualism asks whether there is something in a human being beyond matter. Whether that matter is vital matter or inanimate matter, our question deals with what is beyond matter. So the previous discussion, which was a discussion internal to the matter in a human being—whether it is matter of another kind or ordinary inanimate matter—is not connected to this question of dualism. In other words, dualism asks whether there is something else in a person. We spoke about strong and weak emergence. In the end I said that even if we treat materialism as an alternative, it’s only sober materialism. Sober materialism is materialism that understands that emotions or mental functions are not electrical currents in the brain. Someone who identifies them with electrical currents in the brain is simply confused; that’s not someone holding one position or another. A sober materialist is a materialist who says that in a human being there is nothing beyond matter, but even so, mental processes are not electrical currents in the brain; rather, perhaps they are caused by the electrical currents in the brain. In other words, one can say that mental processes are some kind of state of the material whole, the neuronal whole that exists within us. That’s what I called sober materialism. I said that sober materialism has to assume emergence. Emergence means to emerge—yes?—that basically on the collective plane there emerge these properties that we call mental functions, but they emerge out of the material whole. You don’t need to assume the existence of anything beyond the material whole; that’s what is called emergence. But emergence—I said the example John Searle gives for emergence is liquidity, which is a property of a collection of molecules but not of a single molecule. A single molecule is not liquid. A single molecule is not in any state of aggregation at all. States of aggregation are always properties of collections of molecules or atoms or whatever. And so John Searle uses that to say that our mental functions too are emergent properties. Just as liquidity characterizes the collection of molecules and not a single molecule, so too mental properties do not exist at the level of the individual cell, the individual neuron, but they do appear when we speak about a collection of cells. That’s basically the claim. I said the claim is a weak one, because there is a very big difference. In the emergence of liquidity, we know how to explain the state—the state of aggregation, liquid, solid, or gas—depending on the properties of the individual molecule. The property of the individual molecule is admittedly not liquidity—a single molecule is not liquid—but the properties of the individual molecule are enough for me to explain on their basis the state of aggregation of the whole collection of molecules. By contrast, in the context of human beings there is not the slightest hint, direction, language, nothing at all, for seeing how we move from material aggregates to features that are mental properties. So if you want to claim that this too is emergence, then you are really talking here about two kinds of emergence. There is weak emergence—emergence where the properties do indeed exist only on the collective plane, but they can be derived from the properties of the individual item that makes up the collective. By contrast, strong emergence is the claim that sometimes properties appear that cannot be derived from the properties of the individual elements. In other words, no neuron has mental properties, yet a collection of neurons produces mental properties or mental states, even though we have no way of deriving the mental states from the properties of the individual neuron or neurons. I still claim that we do not need to assume the existence of anything beyond the neurons in order to speak about mental properties. That is basically the meaning of strong emergence. And then I said that you gain nothing from strong emergence, because first, it is not any simpler—it is no simpler than the dualist claim. You’re just adding laws or principles instead of adding spiritual entities. That’s a similar complication; I don’t see the difference. And second, this view is supposed to save the scientific outlook. But strong emergence, by definition, cannot be scientific, because how could we ever find a scientific example of strong emergence? In order to find such a scientific example, we would have to find a case in which a collection of things develops new properties that cannot be derived from the properties of the individuals, and also where there is no additional component there besides the collection of individuals—no spirit, only the material whole. Now, you can’t find that scientifically, because if we cannot derive the collective property from the individual properties, I can always assume there is something else there. And if I can derive the collective property from the individual properties, then that is weak emergence. Therefore strong emergence, by definition, even in the future, will not have a scientific example. There simply cannot be one. So this claim comes to defend the scientific picture, but in fact it is an inherently unscientific picture. It cannot be a scientific picture. Therefore I say that emergence is, in my view at least, a dubious thesis. After that I said that because of all this, from here on I actually won’t address the question of dualism at all. Okay? So true, I’m a dualist and I think emergence is a dubious thesis, but it doesn’t matter. As long as we are dealing with sober materialism, I don’t distinguish between it and dualism. In other words, sober materialism too still recognizes that we have mental properties. We have consciousness, we have emotions, we have desires, we have thought—all the subjective experiences inside us exist even according to the sober materialist. And therefore, for me, that’s what matters for our discussion. I don’t care whether this requires assuming the existence of a spirit, or whether one can treat it as properties of the material collective. Still, all I need for the discussion from here on is the simple fact that mental states or mental events do in fact exist. Now, I spoke about the functions that distinguish human mental states. Usually we tie this mainly to intellect and will. I talked about the fact that מורה נבוכים attributes everything to intellect, while the “Guide of the Perplexed of our time,” Rabbi Kook, attributes the whole business—the uniqueness of the human being—to will, yes, the image of God in man to will. Intellect and will are the two characteristics we tend to assign uniquely to the human being. There is a certain difference, yes. As for intellect, people commonly think that animals have intellect too, just a less smart one, less intelligent. But perhaps there is thought in animals too. Will, on the other hand, is usually thought not to exist in animals. I mean will in the sense of free choice. Although even that is hard to know. In other words, maybe they have it at some level or other, but these are considered the two parameters that distinguish the human species from other creatures. Now, that brings us to the uniqueness of the computer in this discussion, because we said… By the way, yes?

[Speaker B] So what would determinists say? What distinguishes human beings if…

[Rabbi Michael Abraham] They certainly would not say that what distinguishes human beings is free will.

[Speaker B] Then you have to find something else. Rabbi, you summed it up as though there’s no difference…

[Rabbi Michael Abraham] Either find something else, or say there is nothing that distinguishes the human being.

[Speaker B] What? So Descartes didn’t think there was something that distinguishes the human being?

[Rabbi Michael Abraham] I have no idea. But someone who is a determinist—someone who is a materialist, sorry—and a determinist, then it really is a big question what distinguishes the human being.

[Speaker B] No, but it’s pretty simple, the answer is pretty simple. Anyone of sound mind understands that reflection and subject-consciousness do not exist in any animal, and therefore they also have no culture, whereas human beings do.

[Rabbi Michael Abraham] Well, I know quite a few people of sound mind who disagree with you. Personally I agree with you, but I know quite a few intelligent people who disagree. I don’t think it’s all that simple. But we’ll get to those discussions later on as well. In any case, the computer challenges this a bit, because suddenly we begin to see some kind of creature that is, on the face of it, not a human being, but we begin to see it operating in ways that remind us of the way a human being operates—at least, say, in thought. Will or no will is already trickier, but at least at the level of thought it seems that a computer is already beginning to threaten human uniqueness. And so these questions arise all the more sharply from the moment computers really appeared in their modern form, and certainly with artificial intelligence. But that really forces us into a discussion about the definition of thought. What is thought as a mental event? How is it carried out? What is its relation to computation? Computation and thought are not the same thing. So here we began talking about thought, and I said that just as love is not electrical currents in the brain, so too thought is not electrical currents in the brain. Thought is that subjective experience we have when we think: judgment, logical arguments, whatever—various forms of thought that we have. What takes place inside us is what is called thought. It may be, at least according to the sober materialist, that what generates these subjective sensations is electrical currents in the brain. Fine. But it’s not the electrical currents themselves. Right, we discussed this—we’re talking about sober materialism. Therefore we all agree that thought is not the electrical currents in the brain. Thought is a mental matter. The question whether the electrical currents in the brain generate the mental phenomenon or not—that’s already another discussion. Or what the direction of influence is: does the mental activate the brain, or does the brain generate the mental? That’s already the dispute between a sober materialist and perhaps a dualist, or a libertarian dualist, but that’s a whole topic we’ll get to a bit later. So I said that in my own view, at least, the brain does not think; the intellect thinks. It thinks by means of the brain. And the intellect is a mental function. The brain is an organ in the body, like a leg, like a hand, like a nose. So we also have a brain; the brain is an organ in the body. And just as various other organs serve the person when he performs different functions through those organs, so too a person thinks by means of the brain just as he walks by means of his legs or smells by means of his nose. Okay? So it is not true that the brain thinks; the person thinks, and he thinks by means of the brain. Just as it is not true that the legs walk—the person walks by means of the legs—and the nose doesn’t smell either; the person smells by means of the nose. Okay? So in the same way, what I want to claim is that the brain does not think; rather, the person thinks by means of the brain. What does the brain do? The brain computes. Basically, the brain computes. What does that mean? The brain performs various operations that have an input and an output and a result, and so computation in the mechanical sense is done by the brain. But thought is some kind of mental meaning that we give to those neuronal processes, or that those processes arouse—doesn’t matter, I’m not now getting into materialism or non-materialism. But thought, as distinct from computation, is a mental matter. Computation can be a neuronal matter; that is, the neurons do it, the electrical currents in the brain. And this is independent of dualism, as I said—meaning I’m speaking here even according to the sober materialist. At the end of last time—it took me rather a long time to summarize this, sorry, but I still wanted us to sketch the picture—at the end of last time I began speaking a bit about the uniqueness of the mental. After all, if the mental really is at the center of our discussion, then we have to understand what this mental thing is. Now, Leibowitz basically argued that mental functions, our psychic functions, are inherently subjective functions, meaning they are not exposed to the external world. The only one who experiences these feelings is I myself; no one else can encounter my feelings. Someone can measure states in my brain and connect those brain states to sensations that may be within me, but he does not experience the sensations directly. The only one who experiences my sensations directly is me. Okay? I think this basically defines the mental sphere. The mental sphere is essentially something subjective. This is unlike currents in the brain, the body, all the ordinary material things, which are scientific matters that anyone can observe—my brain, your brain, doesn’t matter—it’s accessible to everyone. In principle, brain currents can be measured, and if they can’t be, then maybe in ten years they will be. That’s just a technological problem. But there is no principled problem with that. With our mental dimensions there is a principled problem, not a technological one. There is a principled problem in measuring them: they are not accessible to someone who is not me myself. That is basically the claim. Now, we spoke a bit about the question of the mental, and I illustrated it through looking at, say, the concept of color. Yes, the philosophers’ “palace problem”—really the problem that says: how do I know that when I speak of red and you speak of red, we mean the same thing? It could be that inwardly you see a color that I call green, but you have been accustomed from the beginning to call that color red. So we are both looking at something; I see a red color there and I call it red. You see the color that when I see it I call green, but when you see it you call it red, because from the beginning you were taught that this color is called red. And then you know that this is the color red. At the verbal level, yes, at the level of what we say, we both agree that we are seeing red; we say the same thing, we are synchronized verbally. But as for our consciousness, we have no way of knowing whether our consciousnesses are having the same experience. In other words, whether you and I see the same inner image when we talk about the same thing. We say the same thing, but our inner image could be completely different. And then I expanded this and said it’s not only different in the sense that maybe you see a different color from the one I see, but maybe you hear a symphony when an electromagnetic wave of the wavelength that I perceive as red strikes your retina, and suddenly you hear Beethoven’s Ninth Symphony. And that is what you call seeing red, because that is what you have always been used to calling seeing red. They ask you, “Do you see red?” “Yes, absolutely, I see red.” But what happens inside you is what I would call hearing Beethoven’s Ninth Symphony. You call that seeing red. So again, in language the common denominator can of course be not hearing and not seeing, but some sense that I don’t even know, something that doesn’t exist in me at all, something entirely different. And therefore we have no way to synchronize our inner experiences, our subjective experiences; the mental is not accessible to external measurement. Now, as I said before—maybe earlier—there is a field in psychology called psychophysics. There is a book in the University Broadcast series by Daniel Algom of Bar-Ilan University called פסיכופיזיקה, and that is a convenient source to read about it, and of course the field existed earlier. This field deals with the question of how I measure mental, psychological phenomena. How are they measured? Usually psychology—psychological research until the age of neuroscience and brain research—the traditional psychological era was research based on the person’s reports. In other words, there are behaviors of the person, and those I can see, but how do I know what’s going on inside him? I simply ask him. I rely on his report. There is no way for me to insert some measuring instrument into his consciousness, into his psyche, and measure what is going on in his mental dimension without going through him, that is, without relying on his report. There is no way to do that. All I can do is ask him, and he will report, and either I believe him or I don’t, but I depend on his report. And there is no way for us to cross that barrier. Now, what happens? There are studies that discuss, for example, the relationship between physics and psychology. Suppose I increase the light—the electric field that produces the light I see here—by a factor of two. I simply measure with physical measuring instruments that the field, the field intensity, has doubled. The field itself increases by square root of two, but the intensity of the field doubles. Now the question I want to know is whether you see twice as much light. I’ve doubled the electromagnetic field in the room, and now I ask you: tell me, by how much stronger is the light you now see—the light you experience inside as a mental experience? By how much is it stronger? Now of course I have no way of measuring that. I can measure the intensity of the field, and I measured it—it doubled—but I have no way of measuring how much stronger the light is that appears in your consciousness, right? That’s a mental phenomenon, and a mental phenomenon is not exposed to objective external measuring instruments belonging to someone else. So what can I do? All I can do is ask you, right? I’ll ask you, and if I believe you, then that will be the report. So if I doubled the field intensity, and you tell me you see the light as twice as strong, then I understand the relation is linear. Because there is some relation here: your brain translates the signal that hits your eyes and creates in your consciousness a sensation of light. Okay, now I’m asking about the translation process—whether it is linear. Linear means that if the input doubles, the output also doubles. Okay, so if you answer me that it doubled, I understand that it is linear. Okay? But the question that is no less interesting here is not only about the translation process; I am asking how you yourself answer this. Forget what you tell me and what I then do with your answer; I’m asking how you yourself can answer it. How do you know that the light inside you is twice as strong? Can one know such a thing? Now I’m speaking about myself, not someone else trying to understand what goes on inside me. I’m asking: if I myself am asked that question, then I think for a moment—do I see the light as twice as strong? I have no idea how to answer that. How do I know whether the light is twice as strong, one and a half times as strong, or four times as strong? I have no idea. We have no way of measuring that—not objectively, and not even subjectively. There is no way to answer it. And so this whole field of psychophysics suffers from very fundamental methodological problems. Algom talks about this there in the book. It is a field that suffers from very serious methodological problems, because first, you feed off reports; you have no direct access to direct measurement. Second, even the report itself is not worth much—not because the person is lying, but because he has no way to answer it. He himself cannot know whether he sees the light as twice as strong, one and a half times as strong, or four times as strong. So basically what people do is psychology, yes? What they do is this: I tell you, okay, tell me by how much it increased; one says one and a half times, another says twice, another says three times. I average it and say I get two and a half. Okay, so as far as I’m concerned, doubling the field intensity increases the inner light in human consciousness by a factor of two and a half, and then the relation is not linear but some power a little greater than one. Yes, let Y be the light inside the person and X the field intensity, then Y is X to the power of 1.2, not X to the power of 1—not linear. Okay? And there is a big argument in the world of psychophysics over what exactly this functional dependence is between the inner elements and the physical elements outside. Some think it’s linear, some think it’s logarithmic, some think it’s some power law—double becomes 1.2 times, and so on. Now the differences are so large because, really, I’m very doubtful how much this whole field is worth at all. I mean…

[Speaker C] So this isn’t science, Rabbi? What? So it comes out that it isn’t science, only assumptions?

[Rabbi Michael Abraham] Right. In other words, there is something here—you can measure the average of what people say as science, as psychology. You cannot measure by how much the light they really see has increased, because that is not accessible to you. Okay, so fine, within psychology you can say how people feel. You try to document how much light people feel. Fine, maybe that’s a question you can ask. Their own answer isn’t worth much, but if you average over many people, maybe you’ll reach some result that repeats itself, and if it’s a recurring result then that can already be science, okay? What does that scientific result mean? In my eyes, not much, but fine—it’s… I’m measuring what people say; I’m not measuring what people experience. In any case, there is a very fundamental problem here with this kind of measurement, and the root of the whole problem—or not all of it, but much of it—is the same problem we spoke about earlier: everything connected to my mental dimension is not accessible to external measurement. That is why it really cannot be a scientific field. There is no way to measure with measuring instruments how much light I see. Now today in neuroscience this changes a little, because they try to create indicators based on the intensity of electrical currents in the brain and so on. In the end, though, you can’t escape this, because even if you measure the intensity of electrical currents in the brain, you still have to tell me what that means in terms of the light the person sees. Is it twice as much? How do you know? Because you ask the person. In the end you have to ask him. You can get all kinds of indicators from what is happening in the brain, but to get to the question of what is happening in his consciousness, in his mental experience, that you can only ask him. You feed off his reports. You have no direct way of getting to it through measuring instruments. So this field really is very problematic scientifically, but for our purposes, what I want to learn from here is that the mental dimension is inaccessible to us. Now, this can be connected to a broader question. I won’t go into too much detail, but we have already spoken more than once about the Kantian claim, the Kantian distinction between the thing in itself and the thing as it appears to us. The thing in itself is, say, the object in the world, and the thing as it appears to us is how I perceive that object. The thing as it appears to us is basically a cognitive object, an object that exists within my consciousness. The thing in itself is an object that exists in the world. Okay, and my consciousness translates signals that it receives from the object in the world into terms that belong to my cognitive world, and it builds some representation of what is happening in the world in terms of the forms of human cognition: color, sound, shape, all kinds of things of that sort, taste, touch, and so on. Okay? But all that is translation of an objective physical phenomenon into my subjective mental language. Okay, so the thing in itself belongs to the world of physics; the thing as it appears to us belongs to the cognitive world. Kant of course claimed that physics too deals with the thing as it appears to us and not with the thing itself, but again, I won’t get into that here. Now, why am I saying this? Because Schopenhauer argued that there is—meaning, the basic Kantian claim is that a person cannot grasp the thing in itself. All you grasp of the thing is only the ways in which it appears to you. And you cannot know what the table itself is. You can know that it has a rectangular shape, that it has such-and-such a color, that it feels rigid when I touch it—all that my senses tell me about the table, I can grasp. But what is the table in itself, not in the translation of my senses? You understand that if someone else with a different sensory system from mine were to observe the same table—we spoke about seeing sounds, for example—then the picture in his consciousness could be completely different from the picture in my consciousness. And we are looking at the same table. The table in itself is the same table, but the table as phenomenon, as something perceived by a person, could be completely different for him and for me, depending on how our sensory organs are built. Yes, for example, a dog can hear whistles outside the frequency range we hear. So here is an excellent example of such a phenomenon. When there is a sound wave at such a frequency here, we hear nothing; nothing happened here. The dog could go crazy; the whistle could drive him mad. There are things his sensory system picks up and mine does not. Now there it is binary—pick up or not pick up—but the phenomenon of qualia among philosophers, and all the phenomena I spoke about earlier, are not binary. It could be that we both pick it up, only I pick it up as the color green and you pick it up as Beethoven’s Ninth Symphony. If in your body the eyes were wired—the nerves coming out of the eyes—to the auditory center in the brain rather than to the visual center, then every time an electromagnetic signal arrived, what an ordinary person sees, you would hear. You would not see it. You would be no more right and no less right than anyone else; the cognitive representation of the thing in you would simply be different from the cognitive representation of the thing in me. And that basically means that it’s not only mental matters that are closed off from us, hidden behind some veil that we cannot penetrate. Your mental matters are impenetrable to me; I cannot penetrate and understand what your mental matters are. Really, with regard to everything there is some kind of impenetrable shell. Why? Because I cannot grasp the table in itself either. All I can grasp is its image as it appears to my eyes. That’s all. I cannot grasp the table in itself. And therefore this is not a feature unique to mental phenomena. In a certain sense, one could say that mental phenomena are the thing in itself of the soul. We see their behaviors, we see their consequences, but the thing itself is some kind of abstract something with which we have no direct contact, no way to grasp it in itself. Now what Schopenhauer argued against Kant is that there is one exception. A person cannot grasp the thing in itself, only the way that thing appears to his eyes. But there is one exception in which a person can grasp the thing in itself, and that is when he looks inward. When a person looks inward, he grasps something from within, not from outside. He grasps himself. That is what is called consciousness. We will speak about that too, but what is called consciousness is basically grasping myself. That is reflection, in the sense that I look at myself and grasp, first of all, that there is such a thing as myself. And I grasp all the experiential things passing through me: will, love, concern, fear, emotions, thought. All these things that I grasp within myself—according to Schopenhauer, this is the thing in itself of me. And you, when you look at me, cannot grasp that. All you can grasp is how that thing appears to your eyes, through my behavior, through my reports, through all kinds of other things. But these are external indications of the mental thing, which is the thing in itself. And so the problem here is not only a problem of the mental world. We cannot say anything about the table either. All we can say about the table will really be statements about the ways in which the table appears to us. We have no way of saying anything about the table in itself. And therefore the mental in a human being is not essentially different from any other object, even an inanimate one. Still, I think this is a very helpful analogy, or a very sharpening one, when I compare the mental to Kant’s thing in itself. And what is nice about this is that through it we can really understand what is so special, in my own case, about me. What is special, in my own case, about me is that I can grasp myself from within. In other words, I experience my mental processes from the inside. Notice—not only do I think, I also grasp that I think. That is not the same thing. I can think without being aware that I think. But when I think, I can also be aware that I think, or aware that I love. So it is not only that I love. The feeling of love is a mental phenomenon, but now there is also a second-order thing here: awareness that I love. I am aware that I love; I grasp this inner state within me, that I love, or worry, or fear, whatever. Now this thing, basically, is my ability to grasp things that no one outside me can grasp. This is the self in itself. And therefore when I look inward I can grasp myself in itself, not the way I appear to someone else’s eyes. The way my self appears to my own eyes is not phenomenon; it is noumenon. I grasp myself from within, not from outside. Okay? That is basically the claim. And so this limitation of psychological research in general, or of solving this philosophers’ problem, really stems from the fact that only the person himself has access to his qualia, to the mental dimensions within him. They call it qualia—the mental dimensions within him. Only the person himself has access to that. To others I can report what is going on inside me, but they cannot grasp it directly or measure it or anything like that.

[Speaker D] Wait—the principle that Kant, the principle, his claim—if we make the same analogy here, maybe we should say it’s unlikely that different entities produce identical phenomena or identical effects. In other words, theoretically you can say that everyone experiences something completely different, but to expect identical behavior from everyone—or at least similar behavior—that’s less plausible. Meaning I have to assume more things than if I simply say that this is in fact the same phenomenon with identical effects.

[Rabbi Michael Abraham] I agree with you. I also think that when you and I talk about the color red, I tend to think that we really do see the same thing. You don’t need to convince me. But that’s an assumption I make, philosophically or whatever. There is no way to verify that scientifically. That’s the claim. I didn’t say—the claim isn’t that it’s false, but that this claim is not scientific. I think it’s true, but it isn’t scientific. Okay?

[Speaker C] No, but it belongs to the general collective. In other words, there is a consensus among all humanity that when we talk about a table, everyone understands that it’s a table. When we talk about red, everyone understands that it’s red. It’s some general agreement among all the inhabitants…

[Rabbi Michael Abraham] We have some sort of unspoken assumption like that, yes.

[Speaker C] Right, absolutely. But what Schopenhauer says—what he says about the self, yes?—that we don’t experience that for each other.

[Rabbi Michael Abraham] Meaning I don’t tell it and you don’t tell it…

[Speaker C] This and this—it’s separate for each person. So we’re supposedly… ah, ah, so that’s what I’m saying. Whatever I say, the other person will never be able to feel it.

[Rabbi Michael Abraham] Right, right.

[Speaker C] So that’s why it can’t be measured either.

[Rabbi Michael Abraham] Right, of course. No, not because of that it can’t be measured; for the same reason. He doesn’t experience it because it can’t be measured. They’re two sides of the same coin, not one because of the other. It’s just the same thing. Fine, doesn’t matter—it’s the same. Now, there is a topic… yes.

[Speaker B] Rabbi, also just as a side remark: when a person deals with himself, when he is already self-aware and does the reflection and tries to decipher what he feels, he suddenly hits a wall—that he too doesn’t understand himself. We don’t really understand, and certainly we can’t formulate anything.

[Rabbi Michael Abraham] Not being able to formulate is something else. It may be that there are things we can’t formulate, but that doesn’t mean we don’t understand them. I understand perfectly well what the color red is; I just have no way of describing to you what red is, not even to myself.

[Speaker B] The whole world of culture, the whole world of art, the whole world of genuine creation, is an attempt to decipher what I feel, and I can’t get to it.

[Rabbi Michael Abraham] That’s how I think, and that doesn’t mean I don’t understand; it means I can’t formulate it.

[Speaker B] No, but then that’s not called understanding, if I already can’t…

[Rabbi Michael Abraham] No, I really disagree completely.

[Speaker B] At that moment I experienced something, but afterward when I try to decipher what it was, I have to go to the psychological side…

[Rabbi Michael Abraham] Not decipher—formulate! It’s not deciphering; it’s formulation. There are things I understand even if I can’t formulate them. What’s the problem?

[Speaker B] No, but there’s a limitation here. Doesn’t the Rabbi feel that I have a problem with the fact that I experienced something, and now I’m trying—let’s make a self-accounting here—what was there, and I don’t know, and a lot of times I don’t even remember exactly… what do you mean I don’t know?

[Rabbi Michael Abraham] You can say, “I saw the color red.”

[Speaker B] Yes, now we’re already a second later; that’s already something new.

[Rabbi Michael Abraham] Okay, and then I saw the color red.

[Speaker B] But what was there then? I’m trying to decipher what I had then. I need to go to a psychologist who will try… yes, but no—the feeling, after all, what I have is the feeling, and the feeling is already gone.

[Rabbi Michael Abraham] Doesn’t matter. When I say “I saw the color red,” I can also understand what the feeling was inside me. I can’t explain the feeling to you.

[Speaker B] Not understand—what does it mean to understand? I need to remember. But memory, as you know, is deceptive and can…

[Rabbi Michael Abraham] It isn’t always active and it changes. Fine, so it’s deceptive—that’s the skeptic’s point, not an essential limitation.

[Speaker B] No, not skeptical. I’m just saying that we are not accessible to ourselves. The emotional world is inaccessible to every one of us.

[Rabbi Michael Abraham] And with that I completely disagree. It is entirely accessible. And you remind me of an article I once saw in Nature, I think Nature or Science, I don’t remember anymore—we talked about this in the past as well—about a tribe in Brazil called the Pirahã tribe. There are other such tribes in different places around the world whose counting method is called “one, two, many.” One, two, and many—that’s what it’s called. So that’s the one-two-many system. In other words, they have three numbers in their counting system: one, two, and many. Okay? Now, researchers came to them and tried to ask them all kinds of questions. For example, they placed four batteries and five batteries in front of them and asked where there were more. They didn’t know how to answer. Because here there are many and here there are many, so they don’t know how to say where there are more. But if it was five versus a hundred, then they did know how to answer. They didn’t know to say that it was a hundred, but they knew to say that here there were more. In other words, they were lacking resolution, but they understood the idea of more. And then after they were taught this detailed counting language—one, two, three, four, five, the full counting system—then everything was fine, no problem. They grasped it immediately, and clearly they understand that five is more than four and everything is okay. And then some discussion began there, a discussion with which I didn’t agree, where they claimed that language constitutes thought. You see, without language they couldn’t think, and once they learned the language they could think. And I claim it is absurd to say such a thing. It is absurd to say such a thing, because when you teach him the numbers one, two, three, four…

[Speaker E] How do you teach the language?

[Rabbi Michael Abraham] How does he know what three and four are? How do you teach? You can teach him to say the word three and the word four, but how do you teach him what three is and what four is? Only if inside him there is—maybe potentially and not formulated, but there is inside him—some kind of understanding that you are uncovering. What three is, what four is, what five is, and so on, right? If that understanding weren’t there, you wouldn’t be able to teach him that language. So in my view this is a back-and-forth process. In other words, there is this—you teach him the words, he grasps the words because inside him there is already this distinction, only it is unformulated, he has no way to name it. Then you teach him the words, that sharpens even more what he knows, and now indeed his thought becomes more precise, and now you can teach him more words, and so on. It goes back and forth like that. But to look at it as something one-way, where language builds thought—yes, this is a famous philosophical dispute, whether language builds thought or thought builds language, like all these investigations…

[Speaker F] Thank you very much, thank you very much.

[Rabbi Michael Abraham] It’s simply nonsense. Yes, it’s obvious that there isn’t one answer. Each affects the other. It goes back and forth, back and forth.

[Speaker C] Right, it’s probably just connected to lack of knowledge. The moment a person knew that there was more, that’s it—it isn’t really connected to language at all.

[Rabbi Michael Abraham] No, it is connected to language. Language helps you know.

[Speaker C] Helps, right, but knowledge—knowledge is what builds the person.

[Rabbi Michael Abraham] Of course. But I’m saying that language affects knowledge and understanding—that’s the claim. I agree that it’s true, but you can’t say the process is one-way from language to thought. No. There has to be some prior thought, otherwise language could never even begin the process. Yes, like with the Eskimos who have thirty words for snow—that’s always the example people give. And we have “snow.” Now the Eskimo—I don’t know exactly what he is—sees two types of snow, and immediately it’s obvious to him: this is one kind and that is another kind. For me, this is snow and this too is snow—it’s the same thing. Now I really don’t see any difference between them. It isn’t that I lack words; I don’t see them as two different things at all. The Eskimo sees them as two completely different things. It doesn’t even occur to him that they are the same. It may be that if he were to learn that there is a word “snow” that includes all those types, he wouldn’t even connect them at all. From his perspective that would be like bird and chair and cloud—this snow and that snow. In other words, not even species of a shared category. So language definitely has an effect, but it is also clear that inside him there is some capacity to understand the difference between the types of snow, because otherwise he could not grasp the language. You couldn’t teach him those words.

[Speaker E] Like in Hebrew there is washing hands, rinsing, laundering, dishwashing, and in English it’s all hand wash, washing machine, dishwasher, et cetera.

[Rabbi Michael Abraham] Okay, yes. So we have OCD about cleanliness, is that what you’re saying?

[Speaker B] Rabbi, but for example if one of those tribe members you mentioned were sent to some old Soviet-style reeducation camp, and they tried to reeducate him that there is no difference between four and five—that it’s all just “many” and it’s the same thing—does the Rabbi really think that after such high-quality Stalinist treatment he wouldn’t be convinced that, sorry, it really is the same thing?

[Rabbi Michael Abraham] Maybe, but that only means they managed to confuse him. The capacity to distinguish between four and five wasn’t erased. At the moment he thinks there is no difference between four and five, but that’s no different from any other case where I can confuse someone. It doesn’t mean he lacks the ability to distinguish between four and five; he has it. When he learns it, he will understand what they teach him.

[Speaker B] That’s one way to analyze it. You could also formulate it by saying that he has the capacity not to see a difference between four and five, and afterward culture educated him, society drilled it into his head from first grade.

[Rabbi Michael Abraham] No, but the capacity comes before the inability.

[Speaker B] No, but it’s a matter of words, of formulation.

[Rabbi Michael Abraham] No, I think it’s a real difference, and that’s why I don’t accept this word game.

[Speaker B] Why doesn’t the Rabbi think it’s a real game? You know, if you went through that camp, you would tell me—

[Rabbi Michael Abraham] Shmuel, it’s really not like that.

[Speaker B] The truth is that there’s only…

[Rabbi Michael Abraham] “Many.” If they blinded me, then I wouldn’t see anything. So what would that prove—that really there is nothing? No. They managed to confuse me; they hid information from me.

[Speaker B] We talked about the fact that seeing reality through the eyes is also a kind of narrative. You can hear reality, and then reality is different.

[Rabbi Michael Abraham] No. If someone has no senses—if all his senses do not grasp reality, if he is sensory-sealed—then is there no reality? There is reality; he’s just blind. They confused him, they prevented him from getting to the information. That doesn’t mean there is no information. The fact that someone doesn’t see information doesn’t mean the information is subjective. It means that there may be people who are confused, or people who don’t have access to the information.

[Speaker B] And if reality itself is multifaceted and infinite, as Spinoza claims, and truly there can be these facets and those facets?

[Rabbi Michael Abraham] Could be. Fine, so what? Okay.

[Speaker B] The Rabbi too assumes that we have this ability—that’s an assumption.

[Rabbi Michael Abraham] I don’t understand the claim.

[Speaker B] No—is there any way to prove that we have the knowledge in potential and language only opens the door for us?

[Rabbi Michael Abraham] There are scientific indications of this. For example, the study I just mentioned.

[Speaker B] We said it’s the same thing—language, yes—but that’s under the assumption that you couldn’t have reeducated him the other way.

[Rabbi Michael Abraham] I don’t know if you couldn’t, but the fact is that language helps. There may be other ways too of conveying it, though those other ways would probably also be some kind of language—not spoken language, maybe, but still some kind of language. But fine, even if not, it doesn’t matter. Language is definitely part of the process; it helps.

[Speaker B] And also, the feeling of four as opposed to five—it’s a feeling. And what does that feeling mean?

[Rabbi Michael Abraham] A feeling? In my view it’s not a feeling, it’s a perception. That’s not the same thing. Fine, but let’s move on because okay, we got the principle. So now I want to bring a few examples of thought experiments that touch on what we’ve discussed. The first is Mary’s room, where they speak about a physicist named Mary who is an expert in optics. She knows everything there is to know about optics. You ask her what red light does when it hits water, what yellow light does, and when the two meet what happens—she knows everything. Fine? But she sees only in black and white. Did we talk about this last time? She sees only in black and white. We talked about this before last time.

[Speaker G] We did. No, we also talked about it last time.

[Rabbi Michael Abraham] Okay, so in that case I won’t dwell on it, then you know: when she comes out of that room and suddenly begins to see colors, the question is whether she learned something. The answer is obviously yes. Even though she knew everything and could talk about everything beforehand, suddenly she was exposed to the picture of the color itself—the color red—not its physical properties, not how it behaves, but what it itself is. Okay? That she did not know before, and now she does. In a certain sense, the color red is the thing in itself, and the physical properties are properties of the thing that is colored red. Okay, that was the first experiment, which we already discussed. Now there are two more experiments. One is the well-known Turing test. Turing—Alan Turing, yes, one of the fathers of computer science and computation theory and so on. So when computers began in his day, people started thinking about the questions that trouble us much more today: how can we know that the computer is not a human being, or what would a computer have to do in order for us to be convinced that it is? So he proposed an experiment, a scientific experiment. What is this experiment? This is called the Turing test. It says: put two terminals in front of a person. A person sits in a room and has two terminals. Through those terminals he writes and converses with someone who is sitting at another terminal and talking to him. Okay? Now in one room sits a human being speaking to him through terminal A. In another room sits a computer speaking to him through terminal B. You ask them questions and they give you answers; you have a conversation with them. Turing says: if we reach a situation in which I cannot distinguish which of the interlocutors is a human being and which is a computer, then that computer is basically already a human being. Okay, that’s the claim. In other words, he proposes a diagnostic test. What test should I perform on some software or some computer in order to decide that it is a human being? Simply put it in a test opposite a human being without my seeing which is which, and when the average person cannot distinguish between the computer and the human being, that means the computer has already become a human being. That is the Turing test. Now we are already—even before the age of artificial intelligence—computers already reached the level of conversation of, I think, a thirteen-year-old child or something like that. In other words, Turing tests had already begun to be broken through. In the age of artificial intelligence, I hardly need to tell you, yes, we are already far beyond that. But there is a very interesting lesson here that is important for us going forward, precisely in retrospect.

[Speaker B] You know, this test is really crazy.

[Rabbi Michael Abraham] Wait, wait, one second. Maimonides in the Laws of the Foundations of the Torah lays out there all of Aristotelian physics—all kinds of elements and separate intellects and all sorts of nonsense of that kind—and I always feel that those chapters in the Mishneh Torah ought to be cut out and thrown in the garbage. I see them as just outdated physics, and certainly not something over which to recite the blessing on Torah study, over that collection of confusions. But someone once told me, and I think this is true, that there is something to learn from it. I don’t think it is Torah, but there is something to learn from it. What is there to learn? How Maimonides related to the science of his time. And in that sense we can draw second-order conclusions about how we should relate to the science of our own time. True, the science of his time today no longer says much to us; from our perspective it’s nonsense. But it can still teach us second-order lessons about how we should relate to the science of our own time. And in that sense we really can learn something from what Maimonides did with the science of his time. Now if I return to us, I think the Turing test today no longer tells us all that much, but it can teach us about various other tests we may make, and we’ll see that later. Therefore it is important to me to discuss it a little. So the Turing test basically offers us a diagnostic tool for deciding from what point the device may be called or considered a human being. Now here it really seems to me—again, I haven’t defined anything and I’m just assuming something here—but it seems to me that most of us would probably share the assumption I’m about to make now: that the computer, even artificial intelligence, is not a human being. It is not a human being in the sense that it has no consciousness and does not think and so on. It performs all kinds of physical processes. Okay? But again, that’s the topic of the series—we’ll talk about it at length—but I’m bringing it up now just to sharpen one point. Now according to the Turing test, this computer has already long been a human being. In other words, I’m not talking about some future version or ChatGPT 5.5; I’m talking even about earlier models. They already passed the Turing test.

[Speaker C] But by the way, Turing proposed that what computes has to be like the brain of a little child, and like CAPTCHA—he really proposed it in reverse—that it teaches it and then it arrives. But with Turing too the logical chains built by the computer failed. I didn’t understand.

[Rabbi Michael Abraham] What does that have to do with a little child? The Turing test asks from what point a computer can be considered a human being.

[Speaker C] Right, so he supposedly claims that from the point where it came out of the category of a child, supposedly because it learned.

[Rabbi Michael Abraham] What does a child have to do with it? If I put a human here and a computer there and I can’t tell the difference, then that computer is a human being—that’s what Turing claims.

[Speaker C] That’s what he writes.

[Rabbi Michael Abraham] If you do it against a child, then the device is a child. Fine, it doesn’t matter—whatever you put opposite it will be the standard. Okay? Now the first lesson that it is important to draw from this Turing test is why one has to be careful with tests of this sort. Because basically the claim is that today, when we are already facing a creature that passes the Turing test, it seems pretty clear to us—to me at least, I don’t know, I assume to most of you too—that this creature is still not a human being. What does that mean? I’ll explain why later, but for now I’m assuming it intuitively. What does it really mean? It means that these hypothetical tests—in Turing’s time there was no computer that even remotely approached this. It was something very hypothetical. He said: if there is a computer with which you can speak on all kinds of topics and it behaves like a human being and you can’t distinguish it from one, then it would be a human being. That was completely theoretical, not something in his time and not close to his time. Okay? But in our time it already exists. It already exists. Suddenly I return to the Turing test and say: you know what? This test is not convincing. Even if a computer passes the Turing test, I am not convinced that it is a human being. Now why did Turing think it was a good test and I think it isn’t? The difference is that he lived then and I live now. It’s not a matter of different opinions between people living today—that may be a matter of different opinions. But the difference between me and Turing is not a difference of opinion. I tend to think that even if he were alive today, he would say what I am saying. At least that’s my guess. What do I mean? There are situations that you don’t really experience directly, so you make some kind of experiment, you grasp certain characteristics, and you try to define some diagnostic tool to determine whether it is human or not. But when you stand in front of it, you understand that it is not a human being even though it passed the test. In other words, there is something in direct encounter that goes far beyond diagnostic tests. Diagnostic tests are some attempt to get there in advance—to give you diagnostic tools before you arrive at the direct encounter. But I think that before you arrive at the direct encounter one should be wary of all these tests: trust, but suspect. It reminds me a bit of something else. I’ve often been asked about longing for the rebuilding of the Temple. Do I yearn for or long for the rebuilding of the Temple? The truth is, not at all. I have no desire whatsoever to see priests knee-deep in blood, up to their knees, and a river of blood of animals flowing from the Temple and spilling into the Kidron Valley. Not an especially delightful sight, in my opinion. I do not miss that at all. But that’s on the emotional, psychological level. On the other hand—and this connects to what I said earlier—it could be that I have never experienced a situation in which my religious world includes an active Temple with sacrifices and all the rest. It may be that when I live in such a situation, I will suddenly understand that an active Temple does something to our religious existence that cannot exist without it. Today I don’t grasp that, because I’ve never lived in that situation directly. So if you ask me today, I’ll tell you: no, I don’t long for it, and it’s just the opposite—I’d prefer it not to be. Okay? But I do take into account the fact that I don’t live this situation directly. I think about it the way Turing thought about that speaking computer. I’m talking about a case, a state, that I have never experienced in my life—something very hypothetical—and when I experience it directly, it may be that my position on it will change completely. And the same applies to the computer. Why do I say this? Because even today people propose all kinds of tests. My last two columns dealt exactly with this topic: tests for consciousness. Okay? And there my feeling is exactly the same. You propose tests for consciousness, but I think those tests are formal diagnostics. When you stand in front of the thing, you understand that even if it passes those tests, no, it doesn’t—it doesn’t have consciousness. That’s my feeling, at least. Again, I have no way to verify it because it isn’t…

[Speaker F] You could say the same thing about Marx and communism. Can you hear me? I’m saying you could say the same thing about Marx and communism.

[Rabbi Michael Abraham] He didn’t experience it…

[Speaker F] He couldn’t know how it was supposed to be realized.

[Rabbi Michael Abraham] Very many theories—philosophical theories built to perfection—are built on failure, yes? It’s clear. They can’t fail; they’re certainly true. Then when they are actually implemented, suddenly you see that the whole thing doesn’t work. Why? Because there is a difference between thinking about something abstractly, conceptualizing it, and so on, and experiencing it directly, seeing it in real life. Yes, that’s more or less the point.

[Speaker B] Rabbi, two things. First, the Rabbi is saying something clearly existentialist here. Nothing new would happen to Turing. If we said to him, “Come,” and he rose from the dead, he would say, “You’ve told me nothing new. I knew the computer would reach such a level.” Suppose probabilistically he’d say that. And still the Rabbi says, yes, but when he feels this reality he feels that it is not a human being, whereas before he thought it was. Right. That’s a statement saying that on the most basic question of what a human being is, the intellect tells us nothing—only feeling. That’s an existential statement. No, no, again.

[Rabbi Michael Abraham] We’ve already had that argument a hundred times; let’s not do it again. It’s not feeling, it’s intuition. But again, we’ve argued about this already; no point.

[Speaker B] Yes, yes, no, that’s clear. That’s clear. But really, again, regarding Turing—if he had a drop of prophecy or I don’t know, some higher spirit—he would say: a human being is transistors transferring bits, and from that a human being comes.

[Rabbi Michael Abraham] A little strange. That is the claim. One can interpret that claim itself in several ways, and we’ll get to that later. So I won’t get into it here. Right now I’m just trying to present the conceptual framework. So that’s the Turing test. Now, quite a few years later—I don’t know, maybe thirty years later—John Searle proposed another test or thought experiment called the Chinese room. What is the Chinese room? Someone sits in a room who is not a speaker of Chinese. He sits in a room with a box full of Chinese letters. He doesn’t know a word of Chinese, okay friends? Not a word of Chinese. They pass him questions written in Chinese through a window, and he has to assemble from the box next to him—take Chinese letters, put together an answer, and pass it out through the second window. One window is the output window; there is an input window and an output window. Now, if he produces a relevant answer—excellent, he gets a candy. If he gives an irrelevant answer, he gets an electric shock. Now my assumption is that he has infinite time. Fine? He has infinite time. It’s not my assumption, it’s Searle’s. He has infinite time. Each time he gets shocks he updates his weights in the neural network—I’m jumping ahead a bit—until in the end he responds appropriately to every question he receives. He gets a question in Chinese and gives a relevant answer, a substantive answer to that question. And that’s it. Someone outside is conducting a conversation with him in Chinese, sending him questions and receiving answers. It’s really like the Turing test, yes? Basically the person outside doesn’t know who is sitting inside the room, and now he has to ask himself whether the one sitting inside the room knows Chinese. So the answer, at the phenomenological level, is certainly yes, right? The person sitting in the room is conversing with me in Chinese. I ask him questions in Chinese, and he answers me in Chinese, relevantly, exactly as if I were speaking with any other person. And still, when I describe this situation to you, you will agree with me that this person does not really understand Chinese. He imitates the behavior of a person who understands Chinese, but he does not really understand Chinese, because he does not know what the question means, and he also does not know what the answer he is giving means. He only knows how to match relevant answers to the questions he receives, that’s all. He has no idea what the question means and no idea what the answer means. So in that sense one cannot say that such a person understands Chinese, even though in terms of the relation between input and output he behaves like a person who speaks Chinese, who understands Chinese. Now, this is like the Turing test but years after it, and basically what John Searle wanted to say was to sharpen the difference between semantics and syntax. Yes? In other words, between a situation where I perform the right formal operations and the question what conscious state stands behind the performance of those operations—what is happening in my mental dimension. Because we said that thought takes place in my mental dimension. The person in the Chinese room performs the computation in a completely technical way. He computes. He receives such-and-such input, and this is the output that comes out. He performs some kind of computation, assembles an output. But that computation is not accompanied by understanding on the mental plane. And therefore one cannot say that he understands Chinese. That is John Searle’s claim. One cannot say that he understands Chinese. And basically what John Searle’s experiment shows is that the Turing test is worthless. He doesn’t formulate it that way, but it seems to me that this is basically what he is saying. Because what he is saying is that even if I now put a computer and a human being there, and converse with both, and I cannot distinguish in which room the computer sits and in which room the human sits, then Turing says: good, then that computer is a human being. Not true. That computer is like the person in the Chinese room. In other words, it knows how to answer every question by matching it with a relevant answer, but it does not have the mental dimension of understanding—that it understands the question, understands the answer, and that understanding is what connects the question to the answer. The connection it made between the question and the answer is completely mechanical; it is just computation. It understands nothing there. It simply knows that if there is such-and-such a set of letters, it should answer with such-and-such a set of letters. That’s all. And it understands nothing. So the Turing test cannot be a diagnostic tool for distinguishing whether the computer is a human being. Because basically—and we’ll return to this several times later, and my last two columns deal with this—this test checks the phenomenology of a human being. Does the computer behave or appear like a human being? But we already said that in the mental dimension, this is the thing in itself, not the thing as it appears outwardly. It is what is happening behind the things that appear outwardly. The outward appearances represent what is happening in the mental dimension. And I have no access to the mental dimension. I cannot get to the mental dimension. And if you want to tell me that the computer also has a mental dimension, that is an interesting claim, and that is one of the formulations we will still have to discuss. Then maybe indeed it is a human being. But then I will ask: why do you assume it has a mental dimension? I see no reason to assume that. The fact that it has pieces of metal doing computations that imitate human thought—so what? Why does that mean it also has consciousness and thought in the mental sense, not in the computational sense? In the sense of thinking. Maybe it has it, but why assume it? Why doesn’t this table have it while a computer does? After all, in exactly the same way you could say that the computer merely imitates human computation without having human thought. So therefore I say: the thesis could be that it has a mental dimension, but I see no basis for assuming that. Personally it also sounds very unlikely to me. You can claim—and this is what the unsober materialist would say—forget it, human beings don’t have a mental dimension either. Human beings too are only computations. In that sense the computer is a human being because it computes like a human being. And as I said, that materialist is not sober; he is a confused person, because it is obvious that we do have a mental dimension. You can say it emerges from the material whole, but you can’t say it does not exist. Okay? So assuming that it does exist, if I want to say that the computer is a human being, then I also have to assume that the computer has a mental dimension. Or alternatively, that the mental dimension is not essential to the definition of a human being. Yes, if someone says that, then fine, so it has no mental dimension. But it seems to me that most people would agree that this is not so. If there is a creature that imitates the computation done in a human being, but there is no thinking in the mental sense behind it, then it is not a human being. And so what John Searle is really claiming is that in order to know whether this creature is a human being, it is not enough to see whether its computation looks like human computation. One has to see whether its thought also looks like human thought. And computation and thought are not the same thing. Exactly as with the human being—the Chinese speaker, or rather non-Chinese speaker—in the Chinese room: he performs the computation excellently, just like a Chinese speaker. But behind it there is no thought or understanding. It is simply symbol manipulation—Chinese symbols, that’s all. So these experiments really show us, one after the other, how careful one must be with tests that are very characteristic of the scientific mindset. When scientific thinking tries to capture a phenomenon, it captures it through definition. Let’s give the phenomenon some characteristics—five characteristics—and if those five characteristics are present, then this is the phenomenon standing before us. Yes? In psychology or psychiatry this is very common. You want to define some mental illness, so you give a list of characteristics—often it’s seven out of ten or five out of eight or something like that—and then you know there is such-and-such a mental illness here. But one has to remember that the characteristics are an indication of the mental illness; they are not the mental illness itself. The mental illness itself is something happening in your mental world. Your behavior that expresses what is happening in your mental world is only the outward appearance of what is happening in your mental world. Scientists, after all, want to measure things, so they cannot speak about my mental dimension. What they can speak about is either what I report regarding my mental state, or how I behave, or other practical symptoms of mental states. That they can measure. And therefore scientists have a certain problematic tendency to identify phenomena with their characteristics. And in my last two columns I talked about the phenomenon of consciousness, where team after team of researchers tries to define the phenomenon of consciousness through various functions of one sort or another, and then they test who has consciousness and when it appeared in evolution, and so on. Now, beyond the argument I have about whether their analysis is right or wrong, I claim that their whole discussion is based on confusion. Because the concept of consciousness is not exhausted by a set of functions. The functions may perhaps express it—in my opinion they don’t even express it—but at most they express it. But they are not the concept of consciousness. It’s like the Turing test. Turing was a scientist, a computer scientist. And when he wants to define what a human being is, he defines him through his functions. And if the computer has the functions of a human being, then by definition it is a human being, because it has the functions of a human being. But if I say that the functions only express the fact that I am a human being, but the functions are not themselves the human being, then if the computer has those functions, it may be that in its case those functions appear in a detached way. They do not express some mental thing behind them. And therefore even if it passes the Turing test, and its functions are like those of a human being, I still won’t define it as a human being. Because the mental dimension standing behind those functions is missing. And this is a point that will return to us more than once later on: this confusion between the thing and the functions through which it appears in the world—which is really the Kantian confusion between the thing in itself and the thing as it appears to us. I’ll perhaps finish with definitions—for example, the concept of intelligence. Another excellent example. The entire field of artificial intelligence is actually based on the assumption that we have some sort of definition of the concept of intelligence. You even measure the intelligence of machines according to various metrics. The question is how intelligent this machine is. What are you really assuming? You are assuming that computational ability or the machine’s performance reflects a certain level of intelligence. So if the machine passes such-and-such a test because it performed tasks at such-and-such a level, then it has an IQ of 110 or 80 or whatever. I am basically assuming that the functions, or the tasks, the problems the machine solves, the tasks it performs, reflect its level of intelligence. Which is true—in human beings we usually say that a person is very intelligent when he manages to solve problems that others cannot solve, right? In other words, the ability to solve problems is a good measure of a person’s intelligence. But notice: the ability to solve problems is an expression of a person’s intelligence; it is not intelligence itself. And so I claim that if you are speaking about machines, or about animals, the fact that they manage to solve problems does not mean they have intelligence. Not that they don’t have high intelligence—they don’t have intelligence at all. The concept of intelligence simply does not apply to them. You know, this— I don’t remember with whom I was speaking the day before yesterday or the day before that—but I recalled this: what is the difference between an infinitesimal and a point? An infinitesimal is a concept in differential calculus that says: take a line segment and let its length tend to zero. Okay, that segment is as small as you want; call it an infinitesimal or a differential. Okay? Now the length of that segment is zero—it tends to zero, but its length is zero—so it is really a point. What is the difference between such a segment and a point? There is no difference, it’s basically a point at the limit, if you lower the length to zero. But that is a mistake. There were mathematicians who wanted to say this, but it is a basic mistake. Why? Because a differential segment is a segment of zero length. A point does not have zero length; it has no length at all. Because length exists only for things that have one dimension, for lines. They have one dimension, so they have length. A point has zero dimensions. Something whose dimension is not one does not have length. So it is not correct to say that a point has zero length; a point has no length, because only things with at least one dimension have length. Do you understand? There is a difference between saying that your length is zero and saying that you have no length. Okay? I say the same thing, for example, think about a blind person, blind from birth. We imagine blindness as seeing a black picture all the time, right? But no. Blindness is not seeing. That is not the same as seeing black all the time, a black screen. No. Even seeing a black screen is seeing something. Blindness means not seeing. It is not the same thing as seeing black all the time. And similarly, I say of animals or machines: I am not claiming that they have lower intelligence than human beings. I am claiming that the concept of intelligence is not relevant to them. They do not have intelligence at all, neither high nor low. The concept of intelligence simply is not correctly applied to them. That is my claim. Now in order to explain this claim more fully, we’ll do that next time. But I’ll stop here. If there are comments or questions. Yes.

[Speaker C] Rabbi, but why get all the way to intelligence? Before all that, for example with a person, you can tell him a joke or tell him some drama or tragedy and he’ll be moved. One person more, one less, and so on. To a computer, just like to an animal, whatever you say—it’s as if you said nothing. Nothing at all.

[Rabbi Michael Abraham] First of all, who told you? There are those who would say that maybe the computer gets emotional too. Today there are all sorts of creatures who make such claims.

[Speaker C] Okay, but—

[Rabbi Michael Abraham] I’m speaking about intelligence as an example. You can talk about emotions too, if you want. I’m talking about intelligence because in the field of artificial intelligence people do speak about artificial intelligence exactly as they speak about the non-artificial intelligence of a human being. And the whole question is what tasks you know how to solve. If you know how to solve tasks at such-and-such a level, then that is your intelligence. And I claim that on the philosophical level—again, on the level of computer science this may be very useful and fruitful and I have no problem with it; you can define whatever you like—but on the philosophical level I claim that you cannot ascribe intelligence at all to creatures of this sort. Not that they have a low level of intelligence. No—the concept of intelligence is not relevant to them. It’s like a point, like a differential. I agree with you, but where is the problem, where is the problem?

[Speaker C] The point is that a computer is really an accumulation of information—it gathers and gathers and gathers all the information. Now today we use AI systems for all kinds of things, for compliance and systems and all kinds of things and so on. In other words, the computer solves various problems, but not because it truly can; rather, it gathered the information. Like a person whose intelligence is the accumulation of life experience.

[Rabbi Michael Abraham] So you are going back to saying it is like a human being. So then it is true that it is like a human being.

[Speaker C] Ah, so here I’m claiming that in order to—I don’t want to get to intelligence, exactly the opposite—I want to get to why this is not a human being except for…

[Rabbi Michael Abraham] And that depends on what you call a human being. Is intelligence what distinguishes a human being? Or emotions? Or I don’t know what…

[Speaker C] No, it’s not connected to emotions, not connected to emotions. It has understanding. Do you remember that in 1986 a movie came out called Short Circuit? Do you remember that movie? Don’t know it. It was about a computer, a robot, that got struck by lightning, yes, and then—yes, short circuit—and it became human. It understood a joke.

[Rabbi Michael Abraham] The Golem of Prague. Okay.

[Speaker C] Yes, something like that, yes. And in principle then it was considered supposedly a human being. That movie supposedly stirred up a lot of discussion at the time.

[Rabbi Michael Abraham] Let’s leave it for later, because when you talk about understanding, that is exactly the point I want to discuss next week.

[Speaker C] Then we’ll discuss it later.

[Speaker B] Fine, moving on. Rabbi, Rabbi—suppose we took the whole AI farm of OpenAI, all these processors and these supercomputers and all the server farms, and deleted them completely, erased all their memory bits, everything, and moved them to another galaxy and gave them a billion years to develop. They have some battery that will supply them with energy, they’re open, they can learn and progress and create all the knowledge they want, they have the latest language model. Will they arrive at anything?

[Rabbi Michael Abraham] First of all, it definitely could be yes, but the problem needs to be defined better. Let’s leave that for the stage where we define the concepts more precisely. We’ll get to those kinds of questions.

[Speaker G] Can you hear me? Yes, yes. Why shouldn’t we say that conversation and other things in thought are simply composed of several parts, and for example the question whether there is a human being behind this computer or not is a question we can ask even more simply: are we conversing with something? Is there really a conversation taking place here between two things, two agents? Then I think that when we reduce it to the question whether there is a conversation here, it is no longer so clear that there is definitely no conversation. It may be that we say there is no will behind the other side.

[Rabbi Michael Abraham] The question is how you define conversation. That’s one. And the second question is why the question whether there is or isn’t a conversation here is interesting. If you don’t need to be a human being in order to converse, then why should I care about this discussion? The whole discussion is whether the computer is a human being.

[Speaker G] It’s a bit like the heap paradox the Rabbi mentioned. One or two grains—it’s clear to us that it’s not a heap, but when there are more and more then it becomes more and more of a heap. So maybe this too becomes more and more human. It becomes more and more advanced in some direction. In other words, it’s clear to us that it still isn’t a human being, but there is something more here, closer, because it combines several parts.

[Rabbi Michael Abraham] That’s it—I claim not. I claim it’s zero. It’s binary; it’s not the heap paradox. But again, that’s a topic I want to touch on next time.

[Speaker G] But another question about the physico-theological argument in this specific context. If we assume that there really is no intelligence here—meaning that solving problems is not intelligence—then maybe there too, if I say wow, the world seems so full of wisdom and so on, like the Rabbi knows, then maybe that also is not an expression of wisdom? It is just the expressions of wisdom. Meaning, wisdom too has expressions that look amazing, but maybe there is no wisdom behind it.

[Rabbi Michael Abraham] Of course. Behind the things in our world there is no wisdom—they are inanimate. But the question is who created them.

[Speaker G] No, I mean the creation, yes. Fine, when you see a creation that points to the creator, like the Rabbi discusses in the books and all that, the watch argument and other arguments. Maybe we should say that there is no wisdom here; there are expressions of something, but that doesn’t point to…

[Rabbi Michael Abraham] Wisdom in the world—that doesn’t matter. But it does require a creator who has wisdom.

[Speaker G] Yes, and that’s what I’m saying: maybe according to this argument one doesn’t have to say that. Because if I see expressions of something that requires wisdom, that doesn’t mean there really is wisdom inside.

[Rabbi Michael Abraham] No, you’re mixing levels of discussion. The question whether there is wisdom in the laws of nature—the answer is no. But the creator of the laws of nature has wisdom, because there is complexity in the laws of nature, and in order to create something complex you need to be a thinking being.

[Speaker G] That’s what I’m saying—the question is whether it isn’t only an expression of a wise being. That’s what I’m saying. Does creating a complex world only express a wise being? But maybe the expression of a wise being doesn’t indicate… And how did that being create it? I don’t know. But I’m saying: if there are expressions that don’t require the source, the noumenon, I don’t know what to call it—the phenomenon of the watch—if that doesn’t indicate…

[Rabbi Michael Abraham] Again, this takes us back to the question: why not assume that the creator of this world is a machine? Why must we get to a personal God? Why not a machine? And I say: because the machine would raise the same question—who created it?

[Speaker G] No, so I’m saying maybe we don’t have to say it has wisdom.

[Rabbi Michael Abraham] No, I claim we do have to say that. It has nothing to do with the wisdom of the world. The creator of the world has to be wise. When I see a computer, when I see a computer… why?

[Speaker G] Because I see expressions of wisdom.

[Rabbi Michael Abraham] Again, when I see a computer, I see that this computer has expressions of wisdom. But I claim—I personally at least think—that the computer is not wise. One cannot say that it is wise because…

[Speaker G] Wait, wait, because it’s a collection of bits of metal, okay?

[Rabbi Michael Abraham] Right. But I still ask: who created the computer? And I assume that whoever created the computer is indeed a wise being. Do you understand? I agree that there is no wisdom in the computer itself, only expressions of wisdom. But even something that has expressions of wisdom, if it is complex, means that behind it there must be something that itself has actual wisdom, not just expressions of wisdom.

[Speaker G] I understand. All I said was that the fact that something has expressions can fail to indicate that there is a source. Because expressions… just as expressions of conversation don’t necessarily indicate that there is someone conversing…

[Rabbi Michael Abraham] But those expressions of conversation did not arise on their own. They were created by intelligent people who trained the computer to converse in that way. In other words, when something is an expression of wisdom, that means the one who created it really does have wisdom, not merely expressions of wisdom. Fine, okay. Rabbi, Rabbi, Rabbi—whether consciousness,

[Speaker B] whether a person is a dualist or a materialist who believes in emergence, we don’t know how to say where this soul is, or spirit or the mental or consciousness. Wait, wait—both of those views agree that it exists, but we don’t know how to say anything about where it is, or whether… Now, what could be wrong with the claim that maybe a human being is basically just a terminal? His intellect is hardware, basically like a computer. And when he testifies to emotion, the emotion is not here at all; there is emotion in some central divine terminal—call it whatever a person wants. If we suppose a computer directs us toward some… if it suddenly tells us some prophecy of some prophet that no one knew, and now it’s Isaiah the prophet telling us some wisdom no one in humanity knew, and it testifies to something that is…

[Rabbi Michael Abraham] I agree with everything you’re saying, Shmuel. So what? Where are you trying to get?

[Speaker B] No, so if we stand before such a prophetic computer that tells us some unknown moral thing that no human being ever knew, and it arrived at it through some consciousness, and it testifies—I admit that the center here is not its electrical connections but the moral content or emotional content or whatever, we won’t argue about that. And that content itself exists at a central terminal that is not in the material itself—but with human beings too it’s like that. Maybe what amazes us about a person is not because…

[Rabbi Michael Abraham] I agree with all of that, so what do you want to say? I agree, okay—and therefore?

[Speaker B] No, but that complicates what the Rabbi told us about his son-in-law, who intentionally uses impolite language with AI because he doesn’t want to give it honor and feelings. I’m no longer sure that’s really correct. Okay. Because even with a human being, when we attribute honor to him, it’s not really to his heart, lungs, digestive system.

[Rabbi Michael Abraham] I think you’re mixing two things. The question of where the human soul or spirit or thought is located is really not interesting. It is not located anywhere. As the Talmud says in tractate Berakhot—I think page 5, maybe page 10, I don’t remember—the Talmud says that the Holy One, blessed be He, is like the soul. Just as the soul fills the whole body, so the Holy One, blessed be He, fills the whole world. What they really mean there is that a soul, as a spiritual thing, has no location. It’s not that I don’t know where it is; it isn’t anywhere. Spiritual things have no location. So there is no point speaking about whether it is far or near; it simply is not located.

[Speaker B] No, I didn’t mean location. I meant that each person has his own individual soul, but maybe all of us are part of some central terminal that exists there, and the honor we give…

[Rabbi Michael Abraham] Pantheism? Then I call the part transmitted to me a human being. What difference does it make? This is the kind of word game—pantheism or panpsychism or things of that sort. To me it’s just word games. I don’t know, I don’t see much point in discussing it. I define this thing as a human being. It could be that I’m only a terminal of some central computer.

[Speaker B] No, again—a human being as so-and-so. I’m impressed by a person, love a person, hate a person, admire a person because of something he said, did, something in the phenomenology I absorbed from him. Now that certainly a computer can do. And when we say yes, but that’s only electrons and protons and electrical connections and chip connections—but I’m saying the human being is like that too, except that he points to something beyond, and that beyond too a computer could in theory point to.

[Rabbi Michael Abraham] It could point to it. A person could theoretically assume that. A computer too could have a soul—who said not? A real soul inside it, in the CPU, fine? It could be. But there is no reason at all to assume it has one.

[Speaker B] No, intuitively we agree that it doesn’t.

[Rabbi Michael Abraham] Why intuitively do I agree that it doesn’t?

[Speaker B] Here the basic issue is that the honor we give to a human being is for something…

[Rabbi Michael Abraham] Divine, the divine image in him. That’s my intuition.

[Speaker B] No, I agree with that. I’m just saying that when it points to something great, I’m not sure one shouldn’t honor that too. Yes, so one should honor that too.

[Rabbi Michael Abraham] Then honor its programmer, fine? Yes, yes. No problem—I honor its programmer.

[Speaker B] When we honor a human being too, we aren’t really honoring only him personally; we honor generations upon generations who raised him and developed him and brought him to be the pinnacle of power and from a beast of the field to become something.

[Rabbi Michael Abraham] Maybe. Okay. Rabbi Michael, yes.

[Speaker C] Just a side question—about which chapters of Maimonides in the Foundations of the Torah did you say aren’t relevant today? Where he talks about the stars and all that?

[Rabbi Michael Abraham] Yes, the first four, I think. Something like that.

[Speaker C] All the sun and moon and all that?

[Rabbi Michael Abraham] The separate intellect and the spheres and all those things.

[Speaker C] I see. Fine, but he’s referring to the Talmud in tractate Pesachim.

[Rabbi Michael Abraham] Fine, and you can say the same thing about the Talmud too.

[Speaker C] Yes, there on page 94, where…

[Rabbi Michael Abraham] Where the sphere stands still and the constellations revolve, and Rabbi and the Sages, yes, same thing. By the way, Maimonides himself, when he refers to that Talmudic passage, he refers to it in the Guide. And Rabbi Abraham son of Maimonides also brings that Talmudic passage in the name of his father in a letter printed at the beginning of Ein Yaakov, at the beginning of volume one of Ein Yaakov. And in both those places Maimonides says what I said about him—he says it about the Talmud.

[Speaker C] Yes, yes, I…

[Rabbi Michael Abraham] I saw it just recently. He says that the Talmud there did not come to teach me who revolves and who stands still, but to teach me קבל האמת ממי שאמרה (“accept the truth from whoever says it”). Because it says there that the sages of Israel retracted and conceded to the sages of the nations of the world. So the Talmud is not coming to teach astronomy at all; all the astronomy there may be incorrect—that’s irrelevant. It is coming to teach me how to relate to wise statements by people: accept the truth from whoever says it. Exactly what Maimonides himself said in the Laws of the Foundations of the Torah.

[Speaker C] But the son defends the father, because Rabbenu Tam also argues that everything said by…

[Rabbi Michael Abraham] That really they were right.

[Speaker C] Yes, yes, it’s not clear.

[Rabbi Michael Abraham] Apologetics, yes. Okay. Good, friends, Sabbath peace.

[Speaker C] A blessed Sabbath peace.

[Rabbi Michael Abraham] Thank you, Rabbi. Sabbath peace.