Artificial Intelligence – Lesson 4 – Rabbi Michael Abraham

[Speaker A] It’s all Torah, it’s all Torah.

[Rabbi Michael Abraham] Stay with us, as they say.

[Speaker C] Yes, yes, it’s all Torah in the person, yes.

[Rabbi Michael Abraham] If I were reading Guide for the Perplexed or some other book written in Rashi script, everyone would feel that we were learning Torah. There’s no difference. It’s not… okay. So wait, okay, there’s a recording, fine. I want to start from where we left off in the previous lecture. I spoke a bit about what sets man apart, about mental processes, and now I want to focus us more on intelligence or on thought. I did speak about that, but I spoke about it within the framework of mental processes in general. But now I want to focus more on thought and intelligence, and maybe I’ll begin with the Turing test that was mentioned in the previous lecture. The Turing test basically says: I place a person in front of two computer screens. He’s chatting with someone on the other side through both screens, with two different agents. Behind one screen sits a human being; behind the other screen sits an artificial intelligence or a computer. And Turing said that if the person in the conversation cannot point out which one is a machine and which one is a human being, then we have reached the point where the machine is basically already a human being. Okay, so this is some kind of diagnostic measure: how do we diagnose that some machine has become human.

Now, first of all, there’s a very interesting lesson here, I think — at least if you accept my assumptions — that in Turing’s time, maybe I mentioned this, I don’t really remember, but in Turing’s time this experiment sounded much more convincing than it does today. Today, when we’re already in a situation where there are machines that pass the Turing test — I’m not even sure human beings would pass the machines’ Turing test — but when we already have machines that pass the Turing test, suddenly we, or at least I — again, depending on your outlook — but at least I understand that this test doesn’t really examine whether you’re a human being. Because even when the machine passes this test, I’m still not willing to treat it as a human being.

Now what do I grasp today that Turing, who formulated this test, did not grasp? In other words, what changed? It seems to me this is a general lesson, not only for this issue, and I’ve spoken about it in the past as well. There’s a difference between when I sketch out some thought experiment or some future experiment concerning a situation unfamiliar to me, and a situation in which I actually live through it — meaning, I experience it directly. For Turing, a computer that could pass his test was something off in the future, some imaginary creature. He didn’t really have such a creature before his eyes, and I assume he couldn’t really imagine that such a thing would arrive. It was some sort of theoretical statement. Today we meet these machines, we encounter them directly, we converse with them all the time. And suddenly we understand that this test is missing a very central element if I’m to be willing to recognize that machine as a human being.

And again, there are people who continue with Turing’s view and say yes, if the machines passed then they are indeed human beings in many respects. But it seems to me that the reasonable person — I don’t know, I assume many are like me — doesn’t see it that way. On the contrary: we do not retreat from the definition of the human being, we do not retreat from the claim that a machine is not a human being; rather, we retreat from the claim that the Turing test is a relevant measure. And the reason — or one reason — meaning, the difference between how Turing saw it and how we see it, is the direct encounter with the thing. That is, there are situations that when you do not live them, you do not grasp them directly, but are only fed some description — someone tells you about some situation — then you cannot really feel it and understand it in your fingertips, so to speak. And in that state, something is missing from your grasp. There is something — although the description can be a full description, they can describe to you completely everything that happens there — here in the Turing test everything is described, you have a talking machine, he gave you all the relevant data. But still, something is missing. When you don’t live the situation, something is missing there. You cannot really judge that situation if you don’t live it. Even though the description you have is complete, no details are missing.

And that’s a lesson for many things. One of the examples that drew my attention to this was that people asked me more than once whether I long for, whether I yearn for, the building of the Temple. So I said absolutely not. As far as I’m concerned, let’s say, it’s an apocalypse if the Temple gets built now. There will be priests walking around there up to their knees in blood, and the blood flowing into the Kidron valley, and I don’t know what — things that I am very far from yearning for. But together with that, I try to think about it from a distant perspective. It may be that someone who lives in a situation in which there is a Temple with sacrifices and everything that was there, may suddenly grasp that this gives some added value to our religious world, to our encounter with the Holy One, blessed be He, to our encounter with holiness or something like that, the divine presence that existed in the Temple. Now, all these are very abstract concepts for me. For me, the Temple is some building where they slaughter animals wholesale and the priests walk around there in blood. That is what I see before my eyes when I think of a Temple. But I’ve never experienced it. I’ve never lived in a period where such a thing existed in our world. And I always have to take into account the possibility that this is like Turing. If I were to live and encounter this thing directly, it may be that something would become clear to me that today I simply do not grasp. And again, no piece of information would be missing for me. I learned all the laws of sacrifices and the Temple service, all of Maimonides’ Book of Service, and I know exactly what happens there, all the details, all the laws, what has to be done, who does it — I know everything. I command the material. No new detail would be added if I were to live in that period, for the sake of argument, okay? And still, something in my perception of the situation would be missing. I am not encountering the situation itself. And in that sense, although right now I absolutely do not yearn for the matter, I do write myself a warning note that this may be the result of the fact that it is a distant situation for me. I do not really understand it directly. I know the details that pertain to it, but I do not understand them. I do not grasp it directly. And you cannot judge a situation without grasping it directly. And that is basically the claim. Rabbi?

[Speaker D] Yes. Can I ask a question? Did Turing really think that when I transfer information — I mean, I get that he formulated it so that when I transfer information and get relevant information back, which is maybe how I’d define a conversation, then that counts as my conversing, obviously not with a human being, but with some entity that is conversing with me. Did he really think that thing had consciousness, had feelings? I don’t believe that’s what he thought. Maybe his formulation of conversation was just that I transfer information and get relevant information back — not necessarily exact, but back — in a form we call conversation.

[Rabbi Michael Abraham] Look, I don’t know what was in Turing’s mind. I haven’t read what he himself wrote.

[Speaker D] The question is whether anyone thought it would have emotion, what?

[Rabbi Michael Abraham] It’s not logical — or consciousness, or a

[Speaker D] mental event, what the Rabbi defined, that there are mental events.

[Rabbi Michael Abraham] How do you define the concept of a human being? Again, the question is how you define the concept of a human being. So again, I’m not sure exactly what Turing thought there would be, because the fact is that today people definitely argue that these machines also have awareness, also have consciousness.

[Speaker D] The question is whether it’s like those materialists the Rabbi said are mistaken in their whole understanding of what a mental event is, and that’s what causes the mistake.

[Rabbi Michael Abraham] There are people who don’t even require a mental event in order to define a human being. But there are others who say no — if it has all our functions, then probably, or at least there’s a decent chance, it has mental events too. Quite a few philosophers think that way.

[Speaker D] They would have to say it has that in order to define it as a human being. We’d have to say Turing thought it had mental events.

[Rabbi Michael Abraham] I’m saying, I don’t know what Turing thought, but it doesn’t really matter. For me I’m using him. I’m not going to investigate what he himself thought as a matter of historical research. For me, I deal with the position, not the person.

[Speaker D] I understand, but I just wanted to say that the fact that the Rabbi has the intuition that he sort of made a mistake and said something that today we understand not to be right — that’s because it may be that he was referring to a specific detail, and I don’t think intuition is the only thing needed for this; I can define five criteria that are different here between a human being and the figure you’re talking to.

[Rabbi Michael Abraham] No, no, I’ll get to that, I’ll get to that. Okay. That is one indication of this matter. Last time I also brought up John Searle’s Chinese room thought experiment. And that experiment is often presented as a criticism of the Turing test. Now, in what sense is the Chinese room a criticism of the Turing test? Turing would say yes, this person speaks Chinese — what’s the problem? Let me remind you: a person sits in a room, he does not know Chinese, receives Chinese letters, questions in Chinese characters, answers in Chinese, and he has to take letters, assemble answers from them, and send them out the window. If the answer is okay, all is fine; if not, he gets an electric shock. There is infinite time, so the person learns to match some relevant answer to every question that comes in, in order not to get shocked. Okay. Now the question is: does he know Chinese? John Searle thought it was obvious that he does not. It is obvious that he does not know Chinese. And from his point of view, this Chinese room experiment is a criticism of the Turing test.

Now if Turing — again, Turing as a stand-in, I don’t care about the man himself — thought he was merely describing computational processes but not defining a human being, then John Searle has no disagreement with him at all. Obviously the person sitting in the Chinese room carries out the calculation that leads to the correct answer in Chinese. The question John Searle asked was: is it correct to say that this person knows Chinese? And in that sense he understood himself to be disagreeing with Turing — meaning this is a criticism of Turing. That means he expected Turing to answer that question yes, this person knows Chinese. And that is what he attacks. So the assumption in these discussions anyway is that the Turing test really checks whether you are a human being, not whether you perform thought-like operations like a human being. That is trivial. If you perform those actions, then obviously you perform actions like a human being. What’s the novelty? If you perform actions like a human being, then you perform actions like a human being — what’s new about that? The novelty is that Turing understood that if you perform actions like a human being, then in some sense, at least, you are a human being. Meaning, you probably have other characteristics that human beings have, not just the fact that you perform the actions, because that by itself says nothing. And indeed that is exactly what John Searle attacks. He says: not true. If behind this there is no understanding that stands behind these actions, then you cannot say there is a human being here who knows Chinese.

So again, without resorting right now to what exactly Turing said, this position represented by the Turing test — that’s what I’m dealing with at the moment. And what I basically want to say is that today at least it is pretty clear to me, and I think to most people, that no — this machine, however smart it may be, is not a human being. Now I ask: why? Why is it not a human being? It passes the Turing test. So you can answer in two ways. You can say: because intelligence is not the definition of a human being. In other words, there are other components essential to defining a human being besides intelligence. True, this machine has intelligence like human beings do, but intelligence is not enough. It is not a sufficient condition to define a human being; other things are missing. A second possibility is to say: no, it is not true that this machine has intelligence. Yes, intelligence defines a human being, but it is not true that this machine has intelligence — even though it can do many, many things that we do, and maybe better than we do.

What is the meaning of that? What does it mean that this machine has no intelligence? After all, intelligence is expressed in solving problems, in thinking, in answering questions, and these machines do that. So in what sense do they have no intelligence? Here too, I think I mentioned this in the previous lecture, and I want to sharpen it now. My claim was not that these machines have low intelligence, but that they have no intelligence at all — the concept of intelligence is not relevant to them. If I remember correctly, I gave the example of a blind person, right? I argued that a blind person does not see a black picture; rather, the blind person does not see. And there is a difference between saying you see a black image — when I close my eyes, I see a black image before me. Is a blind person like someone whose eyes are closed? I say no. A person with his eyes closed sees a black image before him. A blind person sees no image at all — it is not that he sees a black image; he does not see. That is something else. It’s a subtle distinction, but I think if you think about it then it’s clear — we are of course talking about someone blind from birth. He simply does not see, not that he sees black. And in that sense I bring this as an analogy for my claim that regarding such a machine, I am not saying it has low intelligence, or that it lacks enough intelligence to be called human. The concept of intelligence does not apply to it. It has no intelligence.

The other example I think I also gave is the difference between a point and a differential or infinitesimal. Right? An infinitesimal is a segment whose length tends to zero, yes? As short as you like.

[Speaker D] A point, yes, that’s a point.

[Rabbi Michael Abraham] What’s the difference? Seemingly a point is also a segment whose length is zero. What’s the difference between a point and an infinitesimal? The difference is that an infinitesimal has length — zero or tending to zero. A point has no length. Not that its length is zero; it has no length. Because in order to have length you need at least one dimension. And a point is a zero-dimensional entity. If you don’t have one dimension, then it is not correct to say your length is zero; rather, you have no length. The concept of length does not apply to you. You understand that an infinitesimal differs in essence from a point, because the infinitesimal is a segment — it has one dimension, not zero. Its length is just as short as you like, tending to zero — you can even call it literally zero for the sake of discussion — but it is still not a point. A point is an entity devoid of length, not one with zero length. Those are two different things. And I used these examples to say that what I claim about machines is not that they have low intelligence, but that they do not have intelligence. The concept of intelligence is not relevant to them. Why not? What?

[Speaker D] Is this what’s called a category mistake?

[Rabbi Michael Abraham] Yes. Now I’ll explain it. For now I’m just warning you — I’ll try to explain it more. But just by way of an example, an interesting example: my brother sent me this just a day or two ago, so I decided to make use of it. Wait. There’s a video here demonstrating a fairly well-known paradox; I once spoke about it in some lecture a few years ago. Are we supposed to see it? Wait. Can you see? …

[Speaker E] Why did the U.S. ban this number in 2001? It sounds insane, but twenty years ago the Motion Picture Association of America was genuinely trying to delete… This is called the staircase paradox. Imagine you’re

[Speaker E] standing

[Rabbi Michael Abraham] at the corner of a perfectly square park. It’s exactly one hundred feet wide and one hundred feet long. And your mission is to get to

[Speaker F] the exactly opposite side in as short a path as possible.

[Speaker G] But you’re stuck in a grid. You can only move north or east.

[Speaker F] Absolutely no diagonal movement. Well first you try walking all the way up the side and then you walk all the way across the top. You travel exactly two hundred feet, obviously. But looking across the park you think to yourself there’s got to be a faster way to do this. I mean the diagonal distance is one hundred forty-one feet roughly, so why do I have to go two hundred? Well you have a good idea: instead of walking one big turn you cut through the middle using a staircase pattern. North, east, north, east. And now you are much closer to that diagonal line. You’re not staying to the edges, it feels like a shortcut. It looks faster. But if you pull out the measuring tape and add up all those short steps, you get two hundred feet. That’s weird. You think maybe I just need to add a few more steps. So you double them. You triple them. You create a jagged path with endless tiny zigzags. At this point your path looks identical to the diagonal line. It lies perfectly on top of it. So logically if it looks like the diagonal line, surely your path will just be one hundred forty-one feet now. Right? Wrong. You see, you didn’t become the diagonal, you just approached it. And no matter how closely you approach the diagonal, unless you become the diagonal, those microscopic steps are still steps. Two-

[Rabbi Michael Abraham] hundred-foot path. You can’t change that. This is a concept called Manhattan distance and in this grid-based world you can represent the… Okay, what is he basically saying? He’s basically saying this: let’s look at this square. This square has side length 100 meters — feet, whatever, let’s speak in meters. 100 meters plus 100 meters. Fine? It’s a square with side 100 meters. The distance from here to here when you go like this is 200 meters, right? And there’s a shorter distance. If we go diagonally, then the distance is of course 100 times the square root of 2, which is 141 meters. Shorter.

So now let’s try to move toward the diagonal. We’ll go half the route this way, half that way, go up, then half that way — that already starts getting closer to the diagonal. What is the length of the route we walked?

[Speaker A] Still 200.

[Rabbi Michael Abraham] Still 200, right? Meaning here we walked 100, plus 100, plus 100, plus 100 — 200. So that’s 200, right? Now I increase it further. Now I make lots and lots of little steps like this all the way up. You remember the picture he had at the end? At the end it really looked like a diagonal from here to here with lots and lots of tiny little steps. What is the length of the route in that case? Still 200. It’s always 200. No matter what the size of the steps is or into how many pieces you divide the side of the square. Right? After all, we divide the side of the square into N pieces and this side into N pieces too, and when N goes to infinity we supposedly get the diagonal, right? But it turns out that even when N goes to infinity, the length of the route remains 200. The limit of this process, the length of the route, is 200, not 141. That’s what he said there: to look like a diagonal and to be a diagonal are not the same thing.

What does that mean, exactly? Why is there such a difference? First of all, anyone not convinced that this really is the result should understand that if I make steps here — this is what he calls Manhattan distance. Why Manhattan distance? Because in Manhattan all the streets are laid out as a grid, at right angles. When you go from place to place, you always go in paths like these. So this is Manhattan distance. And now how do I know it is always 200? Because we always walk horizontally or vertically, right? No angles. I go horizontally or vertically. Now the total horizontal walking I did is this step plus this step plus this step plus this step, and so on. Add up all the horizontal steps and you get 100. Add up the vertical ones and you get 100, right? Sum all the steps — horizontal, vertical, horizontal, vertical — no matter how many stairs you have, overall you walked 200 meters. Okay?

Now I think this is a beautiful illustration. Someone who is not versed in the theory of limits will say that the limit of this process when N goes to infinity, the distance or length of the path, will be 141. Wrong. The limit when N goes to infinity is 200 for the length of the path. Not 141. And the diagonal is not the limit of this process. What is the difference between a diagonal and steps that are as small as you like? Steps whose width and height are each infinitesimal, right? That is the difference. On the diagonal, you are not going horizontally or vertically — you are moving at an angle. Meaning that there are no steps here of zero length. There are no steps at all. If I have steps of zero length, no matter what their length is, the total length of the route always remains 200. Always. Any staircase grid. On the diagonal it is not that it is made of steps of zero length. On the diagonal there are no steps. None. That is exactly the move from infinitesimal to point. And that demonstrates very nicely the difference between zero length and the absence of length. Here you really see it in the calculation. Zero length gives you a path of 200. No length gives you a path of 141. Okay? So you see that the limit of an infinitesimal when it goes to zero is not a point. That is not true. It is a line of zero length. That is not a point. Okay?

All these examples come to show that very often we think that a blind person sees nothing because he sees everything as black. No. A blind person does not see. Not that he sees black. And when I say that machines have no intelligence, people say to me: what do you mean? They can do this task and that task. When I say animals have no intelligence, people say: okay, they’re just less intelligent than you, so they have some intelligence. Fine, but they still have intelligence. But no. I am claiming not that they have a little intelligence; they have no intelligence. That means the concept of intelligence is not relevant to them. It is not the same as saying they have zero intelligence or a little intelligence. No. The concept of intelligence is simply not relevant to them.

[Speaker H] So Dad, if you define intelligence as including emotional intelligence too, then maybe that solves the problem, because emotional intelligence they definitely don’t have.

[Rabbi Michael Abraham] Yes, but first of all there’s no reason for that, and second, it’s not true. They do have emotional intelligence, and artificial intelligence has super emotional intelligence. Today there’s artificial intelligence that can be a psychological therapist. So I don’t think that solves the problem, but that’s not what I mean anyway. It’s not on the plane of emotions at all.

[Speaker E] But why are we connecting it to intelligence, by the way?

[Rabbi Michael Abraham] What? I didn’t understand.

[Speaker E] Why connect this to intelligence?

[Rabbi Michael Abraham] What is “this”? What exactly?

[Speaker E] Your demonstration — when you speak about the computer and when you speak about the blind person, there’s an essential difference here. The blind person doesn’t know what black is. The fact that he doesn’t see is because his brain cannot process the information, he has no ability and has never seen. He doesn’t know what black is, doesn’t know what white is. So the fact that he doesn’t see is built in, obviously. Whereas a computer that accumulates all kinds of information and history and so on — the whole internet belongs to it, so to speak. So intelligence, what we call intelligence, is basically the accumulation of knowledge. Both in a person and in a computer.

[Rabbi Michael Abraham] I understand. My claim is that it is not.

[Speaker E] Meaning, the accumulation of knowledge in a person and the accumulation of knowledge in computing — for you that’s an essential difference.

[Rabbi Michael Abraham] Correct. In a moment I’ll explain. It’s not even accumulation of knowledge. I’m not talking about knowledge accumulation. Soon we’ll speak about the skill of solving problems and so on. Accumulating knowledge is a subtler concept. But I’ll explain. I haven’t explained yet — so far I’ve only declared it. I’ve only sharpened the claim, I haven’t clarified it or justified it yet. But I’ve sharpened what exactly I’m claiming. Now I’ll try to explain it a bit more.

So let’s try to think about this a bit more systematically. When we talk about intelligence, look for example at Wikipedia, the entry on intelligence. First of all it defines the concept. How does it define it? “Intelligence” — in Hebrew “sechel” — is the totality of abilities through which one can solve problems that require thought efficiently. The origin of the word intelligence, etc., has the meaning of choosing, deciding, distinguishing. Accordingly, intelligence is also defined as the ability to understand one thing from another, and as the ability to distinguish between one thing and another. Intelligence is not unique to human beings. Intelligence also exists in animals and artificial intelligence created among machines.” So basically the definition there is a definition of “abilities through which one can solve problems” — the ability to solve problems, okay? And if that is the definition, then animals solve problems, machines solve problems, so both animals and machines have intelligence.

Now I want to challenge that definition, or at least redefine what it means to solve problems. Okay? So let’s think, for example, about animal intelligence. If you saw the example I sent on WhatsApp a few days ago — I wrote there that in preparation for today’s lecture, whoever wants to read it might find it nice.

[Speaker E] About the bees, right?

[Rabbi Michael Abraham] Yes, exactly, about the bees. Here it is. “Bees understand more than we thought.” A new study found that bumblebees managed to solve a complex problem by combining prior knowledge and action planning, which hints at an advanced cognitive ability for goal-directed problem-solving. Okay? Then there are various pictures there. The researchers divided the bees into three groups: there’s some flower there that you need to climb onto a Styrofoam ball to reach, and they divided the bees into groups — those exposed only to the ball, only to the flower, to neither, or to both. Okay? And the claim is that only the bees that were familiar with both the ball and the flower managed to solve the task, to reach the flower by climbing onto the ball.

Then the researchers checked whether this behavior was accidental or whether there was purposeful understanding here. So what happened? They placed two hidden areas where the flower with the nectar, or whatever they wanted from the flower, was hidden behind a partition. After the bees had looked briefly, the possibility of seeing it was blocked. The bee then had to remember that there was a flower behind that partition, go to the ball, roll it, and by means of it reach the flower. Right? So 23 out of 30 bees chose correctly on the first attempt. The bees exercised judgment, remembered the information about where the flower was, and reached the flower. And the researchers said more than that: an important predictor of success was the duration of observation. In other words, if they had looked at the flower for a long time before it was hidden, then they succeeded more. And that means they organize information, or remember it, process the information, and so on.

And so the conclusion is basically that, according to the researchers, the findings join a growing body of research that challenges the assumption that complex cognition exists only in creatures with large brains, and may also affect perceptions related to insect welfare and conservation. That last sentence is interesting. Why does it affect perceptions related to insect welfare? Because the claim is…

[Speaker I] Because if they have intelligence, maybe they also have emotional capacity, and the gap between them and more elevated creatures, maybe human beings, is less obvious. You immediately take it to emotions.

[Rabbi Michael Abraham] You immediately take it to emotions. So no — I… wait, I remember they asked me to do this… You go to emotions, but I’m not intending to go to emotions. What then? Once they have thought and cognitive abilities to choose tools, to solve problems in a goal-directed way, then basically they have certain functions like a human being. Maybe they don’t have emotions, but they do have thought and judgment, planning and execution, capacities for execution. So if so, there is room to begin speaking of them as agents. Meaning, these are creatures about whom we can begin to speak in terms of rights and the prohibition against harming them. They already have some moral status in the world. And why?

[Speaker I] Then surely lions too — they have the ability to plan how to prey on a gazelle, so are you giving them rights?

[Rabbi Michael Abraham] I’m saying that’s the claim in the article. I didn’t say it.

[Speaker I] I’m saying, lions — no need to get to bees. Lions, elephants.

[Rabbi Michael Abraham] True, with bees they found it, and we’d need to define more carefully what exactly they found in them as distinct from others. With lions it’s not certain they found these characteristics of planning in light of data no longer in your possession. Planning on the basis of data that is no longer in front of you, and choosing some way, using a tool to do the thing — there are things here that probably are not found in lions.

[Speaker I] I’ll tell you, it exists in lions. In ants, for example, many dozens of just…

[Rabbi Michael Abraham] This article would not have been published in Science if every lion had it. No, there’s something here — they discovered certain abilities that presumably are not seen elsewhere.

[Speaker I] Because with insects you don’t think of insects as having that level of thought.

[Rabbi Michael Abraham] Never mind, let’s not argue about that, it doesn’t matter. Fine, then lions too. There’s no guardian over lions, as they say. Fine, so…

[Speaker E] Ants have it too. What? Ants have it too.

[Rabbi Michael Abraham] Fine, it doesn’t matter. So anyone who has it — then we can start discussing their rights.

[Speaker E] Science writes about it, presents it as a novelty specifically in such insects.

[Rabbi Michael Abraham] It doesn’t matter to me, I’m not interested right now in what exactly the novelty is, whether with lions or insects. I’m talking about whatever has this, okay? Whether insects, lions, fairies, bricks — doesn’t matter. Whoever has this, it starts raising the possibility that maybe this creature counts as an agent, that it gets some rights, moral standing, that it is forbidden to harm it, that it should be preserved. You know, I already mentioned one of the previous times that today there’s a movement in various places in the world, people who are basically talking about the rights of LLMs, that you may not turn them off without their permission — yes, that basically it’s like killing them. And why? Again, it’s the same form of thinking: once they have intelligence and know how to solve problems and act in a goal-directed way and do things much better than we do in some respects, then fine, you can begin to speak of them as agents. You are beginning to function like human beings, and we need to begin relating to you that way.

So this is a natural outgrowth of the definition of intelligence I read at the beginning. But now I want to take this to absurdity. I want to ask: does water also have intelligence?

[Speaker A] Rabbi, by the way, is this what you explained about the Chazon Ish’s view regarding the labor of building — that you enliven something, you sort of make it alive and dead?

[Rabbi Michael Abraham] Yes, but there I emphasized that I’m not talking about organicity in the biological sense, but rather about a kind of whole in the cybernetic sense — that there is some integrated mechanism functioning as a unified unit. But I didn’t claim that life or consciousness was created there, or anything of that sort. So I return: I want to ask whether water also has intelligence. When I asked people who speak in this language, they all said to me: no, of course not, water has no intelligence. I said: low intelligence, or no intelligence at all? They said: no, no, none at all, no intelligence whatsoever. I said: why? But water knows how to solve very difficult problems that the greatest physicists and mathematicians in the world don’t know how to solve. The Navier-Stokes equations — those are the equations of fluid flow, right? And they too are some kind of approximation, but never mind — they are the equations of fluid flow.

[Speaker E] The physics of water. What? The memory of water?

[Rabbi Michael Abraham] No, no, not memory. I’m talking about solving equations.

[Speaker E] No, they also have chemistry.

[Rabbi Michael Abraham] Fine, I’m talking about solving equations. Let’s focus for a second. What I want to say is this: we have those equations. Physicists and mathematicians do not know how to solve them in any case except the most trivial ones. Some tiny little hole with a drop of water in it — that’s roughly what physicists and mathematicians know how to solve. Nothing beyond that. Now, water flowing over a complicated surface with rises and dips and roughness and all kinds of things like that — there is no chance in the universe that a human being will solve that, even in another thousand years. It can’t be solved. Okay? But water solves it day in and day out — no, moment by moment. All the time, every moment, it solves the Navier-Stokes equations. There it is, flowing, and its flow is the solution to the Navier-Stokes equations for this case. So does water have intelligence — even more than the greatest mathematicians or physicists? It solves problems better than they do. Why not?

[Speaker D] What does it mean to solve? The equations are the description of the flow, no? The equations are the description of the flow, not the solution.

[Rabbi Michael Abraham] They…

[Speaker I] are not

[Rabbi Michael Abraham] solving; it’s just the description of the flow. Fine, but they solve it in the sense of satisfying the law. In what sense are they not solving the problem?

[Speaker I] What’s the difference between them and a computer that solves that problem?

[Rabbi Michael Abraham] The Rabbi didn’t define what it means to solve. The flow is the function that solves the equation. That is a certain definition of a solution. That’s what it is. A problem in mathematics: I give you the Navier-Stokes equation, find me the function that solves it, a differential equation, find me a function that solves it. Here, the water found it. But I solved it and I’m aware that I solved it. Wait, wait — now you’re already bringing in additional dimensions. So far we haven’t spoken about consciousness or anything like that. We’re talking about solving problems, right? That was the definition of intelligence. In my view there is no difference at all between flowing water and an artificial intelligence computer, a superintelligence. No difference whatsoever. They are simply carrying out a mechanical operation that we interpret as a solution to a problem. If that is called intelligence, then water has intelligence too.

When I asked people this question, they said: fine, but water only knows how to solve one kind of problem — it isn’t multi-problem. Artificial intelligence knows how to answer many kinds of questions and solve many kinds of problems. Water only knows how to solve a very, very specific problem. I said: that’s not true, because it knows how to solve the Navier-Stokes equations under many boundary conditions and initial conditions and various bizarre conditions, so it is many problems — but it’s true that they are all problems of solving the Navier-Stokes equations.

[Speaker A] Rabbi, sorry, why don’t you just say water flows by gravity, that’s all? What wisdom does it have? What equations is it solving?

[Rabbi Michael Abraham] I’ll get to that. I’m just doing this in stages. We are at the question stage, not the answer stage. So the claim — what I want to say — I’m taking things to absurdity, okay? I basically want to claim that anyone who attributes intelligence to computers, to artificial intelligence, anyone who attributes intelligence to them, should attribute intelligence to water as well. There is no difference at all between an artificial intelligence machine and water. No difference. The difference is quantitative: water solves only the Navier-Stokes equations, and artificial intelligence solves many kinds of problems. Fine, so maybe water has little intelligence, but still it has intelligence.

[Speaker D] Reaching the result doesn’t mean I made the path.

[Rabbi Michael Abraham] But the computer didn’t make the path either. It also just reached the result.

[Speaker D] Yes, yes, that’s what I meant. The principle in many things the Rabbi has spoken about is that having the result at the end doesn’t mean I made the path.

[Rabbi Michael Abraham] Again, all of you are jumping ahead to where I’m going, but since for so many people in this area it isn’t clear, I’m doing it pedagogically. I’m doing it in stages. In general I agree with everything being said here — that’s exactly where I’m heading.

So what I want to say is that water has no intelligence — not low intelligence because it can solve only one kind of problem. If you measure by how many problems you know how to solve, then water has intelligence, just low intelligence. A computer once had higher intelligence, and today’s artificial intelligence has even higher intelligence. Okay? That’s how they measure it, by the way, in the field of artificial intelligence: they measure machine intelligence. And their definition of intelligence is exactly this one. Therefore, in my view, according to the accepted definition in the field of artificial intelligence, water has intelligence. You could even give it an IQ test and attribute some level of intelligence to water. It will know how to solve only the Navier-Stokes questions. Fine, then maybe that gives it an IQ of ten. Okay? So in that sense it has low intelligence. My claim is that something essential is missing there. My claim is that water does not have low intelligence — water has no intelligence at all.

And now I return from the water, which is just an extreme example, back to machines, and maybe to animals as well. About animals one can hesitate, but maybe also about animals. But about machines, certainly. I claim there is no difference: machines have no intelligence either. Of course, you can use some alternative concept of intelligence metaphorically in order to evaluate the problem-solving ability of a machine. Fine. In the field of artificial intelligence it is certainly correct and useful to define different metrics for machine capabilities. But I would not call those metrics intelligence. Though fine, that’s just terminology, because maybe they parallel human intelligence in some technical sense. But this is not intelligence. It parallels it in some technical way, but it is not intelligence.

Now again, I assume some people in AI would agree with what I’m saying here; they would just say, fine, but for our purposes it is very useful and productive to define intelligence that way, because then you can measure things, make progress, compare one machine to another, and so on — and that is the business of the field. That’s perfectly fine. I’m only saying: if this is a methodological assumption, excellent. The moment you turn it into a claim about the world — and I talked about this when we discussed vitalism, if you remember, how people turn methodological assumptions into claims about the world — well, here too it’s the same thing. People have a tendency to take a methodological claim or methodological definition and turn it into a claim about the world. That is exactly where the concluding sentence of the article we just read comes from. The sentence that says: wait a second, if bees can solve problems, then they have rights and we need to rethink whether one is allowed to kill a bee. Okay? What is that leap? You can say: well, because I think maybe they really also have judgment and the whole package that human beings have. Maybe. But to my mind that suspicion is very weak. It is far more plausible to me — not necessarily, but far more plausible — that the bee is some kind of machine, and that it knows how to solve problems in a goal-directed way just as artificial intelligence does, and therefore I would still say that it has no intelligence, just as water has none and artificial intelligence has none. So I have no problem if you want to measure the bee’s abilities and say whether it is equivalent to Opus 4.8 or GPT 5.5 in terms of capabilities.

But the moment you jump to the philosophical level and say: wait, if so there are philosophical consequences, bees have rights, one may not kill them — there you have already made a leap. Because there you turned a methodological definition of intelligence into a philosophical or essential definition of intelligence. You are basically saying: this is an intelligent creature. That is a leap I am not willing to accept. I’m not sure bees aren’t intelligent, but I don’t think the fact that they have problem-solving abilities means that they are.

I’m phrasing it cautiously, because really you can’t know. It may be that animals really do have what we have, only perhaps at a lower level. The same is true of bird navigation, of physical processes happening inside a molecule, or an atom, or an inanimate rock. There are super-complex processes there. The Schrödinger equation for electrons inside an inanimate rock is something insane — no physicist in the universe can even begin to come close to solving it. But the rock solves it all the time. Water is just an example, but I could ask this about everything in the world. Its complexity is so enormous that we really cannot solve the true equations of anything. Science always deals with cases that are very, very theoretical, purified, simple cases — never real-life cases. Science never deals with real-life cases. It draws conclusions about real-life cases, but solutions in the full mathematical sense — that is never a real-life problem. I don’t think anybody has ever solved a real-life problem in the scientific world. There are various approximations, but that’s… Therefore I say that… Rabbi…

[Speaker I] Yes… I still haven’t understood how the Rabbi mixes together intelligence as solving problems and the definition of a human being. If we take a person — say theoretically, and there are such people at certain levels — a person who in terms of problem-solving is a complete genius, 180 IQ, but somehow he has a problem, as the Rabbi sometimes likes to describe it in those black series, of total emotional flattening, no affect at all. He feels nothing, suffers nothing, at the extreme. The center of suffering in his brain was damaged, he feels nothing, but if you throw him a problem he solves the most complex one in the world, in mathematics, in physics, in computers, in everything. Okay. His definition as a human being is very, very complicated.

[Rabbi Michael Abraham] Fine, so we’ve returned to the emotional plane. I’m not interested in the emotional plane. What interests me is his problem-solving.

[Speaker I] So what remains? Then a computer? Here, there’s a computer, it has a chip-brain, it solves problems — what’s missing?

[Rabbi Michael Abraham] Here is exactly the point we’re arguing about, and now I’m coming to it. No. What remains after removing emotions is not a computer. Absolutely not. A major mistake. And that is exactly what I’m asking now: what is missing in the person in the Chinese room, in the computer, in the water? Why not define this as intelligence? I claim that essentially they are not really solving the problems. Not because they have no emotions — that’s not it. It is simply not true that they know how to solve problems. That’s a mistake. They don’t know how to solve any problem. Why? Because the ability to solve problems — and I am willing to accept that the ability to solve problems is the criterion for intelligence — but what does it mean to solve a problem? To flow through a given topographical layout is not called solving a problem.

What is missing there? I think two things are missing. First, if that thing is a deterministic process, then it cannot count as solving a problem. Right?

[Speaker I] Why? Why?

[Rabbi Michael Abraham] Because if water’s behavior simply follows from its nature, if the way it flows follows deterministically from its character, from the laws of nature — it’s a deterministic matter. Then so is the computer… wait a second, Shimon, wait a second… In such a place I do not see that thing as solving a problem. Think, for example, of a computer. Suppose I fed it some program, and that program, say, solves a quadratic equation. The computer does not know how to solve a quadratic equation; the software doesn’t know either. The one who wrote the software solved the quadratic equation. The one who wrote the software is the one who actually solved the problem. The computer merely carries out in practice the plan conceived by the programmer, and the programmer is the one worthy of being called the problem-solver, not the computer and not the software. Therefore I say…

[Speaker I] Why? Why?

[Rabbi Michael Abraham] Wait… If water supposedly solves the Navier-Stokes equations — no, that’s not true, it doesn’t solve the Navier-Stokes equations. Whoever created the world in such a way that water can find its path in such a complex situation is the one who actually created a world whose laws of nature are described by the Navier-Stokes equations, and the world, so to speak, solves the equation. The one who solved the equation is the programmer or the creator of that situation. Because solving a problem in a way you were programmed to do is like an arrow: we wouldn’t say the arrow has intelligence and knows how to fly, while I don’t know how to fly. No. The arrow doesn’t know how to fly; I shoot it. The ability to send the arrow flying is my ability, not the arrow’s. The ability of water to flow through a given topographical layout is not the water’s ability. It simply follows deterministically from the nature of water or from the laws of physics governing water’s movement. And therefore you cannot attribute intelligence to water.

Intelligence is attributed to a person when the person chooses which tool to use in order to solve the problem and then uses the tool. Choosing which tool to use is the solving of the problem, not the actual execution of the solution. Choosing which tool is the right one to use — that is the solution to the problem.

Now note: even in the case of a human being, if someone says this process occurs deterministically, then I will say that a human being does not have intelligence either. Why?

[Speaker E] So then…

[Rabbi Michael Abraham] because his nature performs a calculation like a computer and arrives at the solution to the problem. For me, the concept of intelligence is a concept that describes the way you solve the problem, not the execution of the solution. You arrive at the solution, you plan the solving of the problem. That thing involves what I call judgment. Water has no judgment and a machine has no judgment. A machine calculates; it does not think. That is a very big difference.

A human being performs the calculation after he has thought how to do the calculation. He chooses a tool. Suppose I have a formula for a quadratic equation. So I’m given a quadratic equation to solve, and I choose this tool, the quadratic formula, and now I apply the tool and solve the equation. Am I an intelligent person? Yes, because I chose the right tool. But you didn’t have the possibility of choosing אחרת. A computer that is fed that tool and solves the problem mechanically is not an intelligent entity. You can solve a quadratic equation also by graphing it and finding where it intersects, without using the algebraic equation. That too is a tool, no problem. But you choose the relevant tool to solve the equation. The moment you chose, you exercised judgment. You could have chosen this tool, you could have chosen another tool; you decided to use this one. That is called intelligence.

[Speaker I] And what is

[Rabbi Michael Abraham] judgment?

[Speaker I] judgment, if deterministically you couldn’t choose anything else? But you didn’t really have free

[Rabbi Michael Abraham] choice.

[Speaker I] Of course you did. What — if I asked the Rabbi, two plus two is four, what options are open to the Rabbi to choose?

[Rabbi Michael Abraham] No, because there are things I already know from previous experience, never mind. But the first time I did this I certainly chose a solution tool. This is what Rabbi Dessler calls the “window of choice.” After we accumulate skills, there are already things we do automatically. Fine, and then indeed we will need to examine what we do in places where we don’t have an automatic solution. When you get to graduate school and now need to do research — not apply material you already learned to another problem of the same type, but do research and choose a path for solving the problem — that, at least in my view and worldview, is not a deterministic process. You need to decide which method to choose, which tools, how to proceed. You think what makes more sense, what makes less sense, and then you also display abilities in using those tools. Even operating the tools requires some kind of ability. And those are not things that emerge deterministically. And only because of that can we attribute intelligence to it.

[Speaker I] So is free choice the concept of intelligence?

[Rabbi Michael Abraham] I call it judgment. Free choice belongs to the value sphere; judgment belongs to the intellectual sphere.

[Speaker I] Maybe intellectual free choice — that’s only…

[Rabbi Michael Abraham] That’s just terminological, never mind. I prefer to call it judgment. It’s a choice, never mind. For me “choice” relates more to the value sphere, and here it’s judgment, but it’s the same idea. You have the freedom to choose between alternatives and you decide — not causally, but purposively. You know that this tool will lead you to your goal. How do you know? You exercise judgment, intelligence, and then say: good, I’ll use this tool. Sometimes I’m mistaken and I discover that it was wrong, so I choose another tool. This process of trying, trial and error, choosing tools, drawing lessons, learning from trial and error — this process is a process of intelligence if it is not done deterministically. Because a machine too draws lessons and learns from examples it encounters. Except that in a machine it happens deterministically.

[Speaker I] So here, Rabbi, let’s say we reached a utopian reality where the Rabbi knows everything it is possible to know about reality — about the world, physics, chemistry, mathematics. The Rabbi has reached the end of days where everything is known to him. Therefore he no longer has any possibility of choosing to exercise judgment about what to know, because everything is known. At that moment the Rabbi stopped being a human being.

[Rabbi Michael Abraham] I didn’t stop being a human being.

[Speaker I] You became a computer.

[Rabbi Michael Abraham] I didn’t stop being a human being.

[Speaker I] He has no choice, nothing to do, no judgment, nothing to exercise judgment about because everything is known.

[Rabbi Michael Abraham] A person who is never in situations of moral dilemma is not a human being? If he never encounters a moral dilemma, he’s not a human being?

[Speaker I] I don’t define a human being that way. I think a human being is someone who suffers.

[Rabbi Michael Abraham] You’re mistaken, badly mistaken. Because there is a difference between your ability to solve problems and the question of whether you actually make use of it. The fact that you don’t need to use it because you encountered no problem does not cancel anything in you. You have the ability to solve problems. The fact that in practice you won’t need this ability because all your problems are already solved — okay, so what? But the ability to solve problems is what is called intelligence.

[Speaker I] He has no choice at all, Yossi, everything became deterministic. And that’s a potential ability of…

[Rabbi Michael Abraham] No, no. You simply do not encounter problems.

[Speaker I] Rabbi, that bee also has a potential ability if we let evolution work on it — why, if I let evolution work on it for a trillion years it will become the Rabbi.

[Rabbi Michael Abraham] Evolution works on it — that is exactly what I’m saying. That’s the point. That is not its intelligence. It is a mechanical deterministic process that simply happens. There is no one there making decisions about what to do and what not to do.

[Speaker I] But the Rabbi agrees with me — the Rabbi supports evolution. Once we were a bee, we were a monkey, we were I don’t know what, an amoeba. And the amoeba became me.

[Rabbi Michael Abraham] When we were a bee, we didn’t have intelligence. Now we do. Why is that relevant?

[Speaker I] It’s not…

[Rabbi Michael Abraham] No, because you’re talking about potential, because you’re already admitting that there is something here that is…

[Speaker I] No, but there is no potential.

[Rabbi Michael Abraham] Not potential. I have the ability to solve problems, not potential. Actually. But there are no more problems. You spoke of a utopia where there are no more problems. I’m not using it. I have the ability, but I’m not using it. What does that have to do with anything? So what? If I can run a hundred meters in eight seconds, I run very fast. I have the ability to run very fast — not potentially, but actually. I just won’t use it because I don’t encounter a problem or need to use it. So what?

[Speaker I] To define a person as a potential problem-solver and that’s the whole story?

[Rabbi Michael Abraham] That seems to me a very thin definition. Not potential — again, you’re going back to potential. No, it’s not potential. It’s an existing ability, not potential. The ability exists; it’s not potential. But I’m willing to accept… still, it’s not potential here, there’s no need to call it potential.

[Speaker E] Listen, but ninety percent of us are not living in problems that we solve every day. We invent things for ourselves to do and we live as human beings because we read books and study and advance and exercise judgment every single day. The Rabbi is one hundred percent right here. We’re not…

[Speaker I] I don’t think that’s what defines us as human beings, really

[Speaker E] really not.

[Rabbi Michael Abraham] I don’t understand the question. We exercise judgment, don’t exercise judgment — I don’t care. I don’t care how much judgment we exercise. I’m talking about the ability to exercise judgment, that’s all.

[Speaker I] I don’t care how many times a day I use it, that’s not relevant. But suppose in principle we have no problems. A baby, a baby has no ability. A baby has no ability to exercise judgment and solve Schrödinger’s equation. But it has the ability to learn and to become, potentially… so an amoeba also has that ability if given enough time.

[Rabbi Michael Abraham] But Shmuel, a baby truly has that ability only potentially, and therefore indeed a baby does not possess judgment, period. What’s the question?

[Speaker I] And still we think that he is…

[Rabbi Michael Abraham] a human being. We feel that he is a human being. Correct. What does that have to do with this? That’s another discussion. I haven’t yet spoken about what a human being is; I’m talking about what intelligence is. A baby has no intelligence. None. Because it does not possess the ability to solve problems. It has the potential to develop such an ability, true. Now we can start discussing what is required in order to define you as a human being. At the moment I’m defining what intelligence is, not what a human being is. A rational creature.

[Speaker D] Yes, exactly.

[Rabbi Michael Abraham] “A rational creature” — I think that’s what you said, Shmuel.

[Speaker D] Rabbi, someone who is addicted, drugged, whatever, and now wants to get his fix, his drink, his gambling. He is now controlled by this, and now he thinks: how do I get the money or whatever for the gambling…

[Rabbi Michael Abraham] But in that thinking he is exercising judgment, so he is exercising judgment.

[Speaker D] But no, in my opinion there is no choice here; it could be deterministic.

[Rabbi Michael Abraham] No, he has no choice in the sense that he has no choice not to consume the drug. He has no choice in the moral sense.

[Speaker D] But the way he gets it is not deterministic.

[Rabbi Michael Abraham] He exercises judgment as to the most efficient way to get the drug. That is not deterministic — he chose it.

[Speaker D] Yes, but the judgment — I mean, he just does the best thing he understands he can do.

[Rabbi Michael Abraham] But he uses judgment in order to understand what is best.

[Speaker D] Yes, but that judgment will simply happen automatically. He won’t have another judgment.

[Rabbi Michael Abraham] Oy. The fact that you choose your goals in a way that is forced on you is irrelevant. Still, in order to achieve that goal he exercises judgment. There are two people, both want to achieve the maximum amount of money. But one does it better than the other. So what, the first one is not more intelligent than the second because both want money? You’re talking about the ways of realizing your goal. The ways of realizing the goal — that is judgment.

[Speaker J] Maybe there’s an additional component that can be added to the Rabbi’s definition. The Rabbi said that a person with intelligence exercises judgment in order to solve problems. I think there’s another component: he’s also the one who determines what the problem is. He defines the problem. When water flows, it doesn’t flow because it decided it has some problem according to the equations; it flows because it flows. But intelligence…

[Rabbi Michael Abraham] You’re right, but that is part of the same feature I’m talking about — water decides nothing. Neither how to solve the problem nor what the problem is.

[Speaker J] It doesn’t decide because it has no choice.

[Rabbi Michael Abraham] It has no judgment.

[Speaker J] I’m adding this component because judgment in solving problems skips the stage in which the intelligent being sets the problem, defines it, and knows how to exercise judgment.

[Rabbi Michael Abraham] Correct, I’m saying it doesn’t materially change the picture. It’s like the problem of the addict that was raised here before.

[Speaker J] No, on the contrary, it solves exactly the problem of the addict.

[Rabbi Michael Abraham] No, it doesn’t solve the problem.

[Speaker J] No, the addict knows he must obtain the drug — that is his problem — and now he…

[Rabbi Michael Abraham] exercises judgment and intelligence.

[Speaker J] Correct, I agree.

[Rabbi Michael Abraham] even though he does not set his goals before himself.

[Speaker J] He has to set the problem, because judgment…

[Rabbi Michael Abraham] No, the problem exists physiologically. He did not choose the drug.

[Speaker J] Doesn’t matter. When he goes to obtain the drug, he gets up in the morning and says, I have a problem: how do I get the drug, because my body needs it.

[Rabbi Michael Abraham] He chooses the most efficient way to get the drug. Fine, but he chooses the way. The question whether you choose your goals or don’t choose your goals is a special case of the question whether you choose something. Whether you choose the goals or choose the ways to realize the goals doesn’t matter to me. As long as you choose, or exercise judgment, that is what I call an intelligent entity. Okay? Therefore I don’t think it matters essentially whether you choose the goals or choose the ways to reach the goals — either one is fine. On the contrary, many times choosing the goals doesn’t necessarily indicate intelligence. You are satisfying some need. So in my view it is often the ways of solving, of reaching your goals, that express more intelligence.

But there is another component missing here, I think, and it’s related — the component of consciousness. My inclination is to think that consciousness is also essential here. I’ll explain what I mean. There were experiments by Benjamin Libet, an American neurologist, who tried to test whether a person has free will. He sat people in front of a clock with a hand running around, and beside them on the table was a button. He told them: “Whenever you decide to press the button, press it — your choice. But I only ask that you keep looking at the running clock and tell me at what moment you made the decision in your head.” Okay?

Now what happens? There were electrodes attached to the head, and they measured an electrical potential that arises before the person presses the button. This was already known before Libet — there is an electrical potential that arises before the person presses the button. The readiness potential, RP. Okay? So we can measure this with EEG and see there is a readiness potential, and after that the person presses the button. The interesting question is: where is the decision to press the button located? We know when the potential arises and we know when he presses the button, but we do not know when he decided to press the button. Again, this is related to mental events, which we have no way to know except by the person’s own report; we discussed this in previous sessions. So we ask him: “Look at the hand moving there and tell us where the hand was at the moment you made the decision.”

Now what happens? We compare his report to the moment the potential arose in his brain — he doesn’t know when that happens, we measure it for him. Why is that important? Because if he made the decision after the potential had already been measured, then that means that decision did not determine that he would press the button. After all, even before he made the decision, we already saw that he was going to press, from the potential that had arisen there. So therefore he has an illusion that he made a decision, but in truth the event he calls a decision was caused by the electrical potential that arose in him in some mechanistic, deterministic way. By contrast, if the moment of decision preceded the arising of the potential — then apparently there is free will and the decision is what aroused the potential that eventually caused the action. But the decision started the chain.

Therefore what is important in the Libet experiment is where the moment of decision is located relative to the moment the potential arises. Okay? Now to his surprise — Benjamin Libet believed in free will — to his surprise he discovered that the potential arises before the decision is made. Which means that free will is an illusion. We basically know he will press before he decides. It is forced upon him. He just has some illusion that he is deciding, but in fact we knew it before he “decided.”

One of the objections to Libet’s conclusion — there are many objections, I wrote… wait, wait, let me finish. There’s a chapter in my book The Science of Freedom where I discuss this. One objection to the experiment says: maybe a person does have free will, only it is unconscious free will. When he becomes aware that he decided, that indeed is after the readiness potential. But before the potential arose, he had an unconscious decision, and it was free. Therefore the fact that he reports the moment of decision after the potential does not mean he lacks free will. Maybe he has free will, only not conscious. He chose freely — meaning non-deterministically, right, that’s what “free” means — but it was not conscious.

Now why do I not accept that objection? Not because it is wrong — maybe it is right — but because philosophically it doesn’t interest me. Why not? Because if you decide freely, non-deterministically, but not consciously, then that is not really judgment. Judgment is only something done consciously. You understand, you weigh, now you decide — now I decide to press the button. That is called exercising judgment. But if some unconscious processes are occurring in my brain, and they are free, non-deterministic, let’s say for the sake of discussion, okay, but I am not aware of them, then I cannot say that this is an action I performed, right? It is not an action in which I acted as an agent, exercising judgment. Even though something non-deterministic happened here — supposedly a free decision, supposedly judgment.

Therefore what I want to claim is that for the person, or that entity, that creature, to count as an intelligent entity, it is required first, that it not be deterministic; and second, that it be conscious. Therefore, when I want to regard an animal, a machine, water, whatever you want, as an intelligent entity, I basically need to assume — whether I’m right or not — but it requires the assumption that this does not happen deterministically and that it happens consciously. That the entity is conscious of the judgment, exercises judgment, and decides consciously which tool to use or which path to take. Okay? If these things do not exist — and again, I’m not saying we can know whether they exist or not; we cannot know whether machines or animals have them — but I’m saying: anyone who claims these things have intelligence also has to claim that these things have consciousness and choice, free will that is non-deterministic. Because without that, it is a mechanical action. And even if it is not mechanical, it is not done out of decision or judgment by this creature. And that is not an intelligent action.

Just as on the moral plane, if an action is not done in this way, I will not call it free will, so on the intellectual plane, if it is not done in this way, I will not call it rational action or intelligent action. Rabbi? Yes.

[Speaker E] In that experiment he did about free will, wasn’t there a person in whom the will supposedly kicked in and then he regretted it and didn’t press?

[Rabbi Michael Abraham] The question of veto. So Libet argued yes — a person can impose a veto. Meaning there were situations in which the readiness potential arose and yet he did not perform the action. There were no situations where an action was performed and beforehand there was no readiness potential — that did not happen. But there were situations where the readiness potential arose and the person vetoed it. Therefore Libet argued that his experiment does not prove there is no free will, because of the fact that you can impose a veto. But again, I’m not getting into all these matters — you can read in my book The Science of Freedom. Here I brought it only as an example to explain why, in my view, consciousness is also important and not only non-mechanicality. In other words, an intelligent action is an action of solving problems through conscious judgment. That is the definition I am proposing.

[Speaker D] So according to this, all the materialists are — according to materialism there is no intelligence.

[Rabbi Michael Abraham] The determinists, not the materialists.

[Speaker D] Yes, determinists — why not materialists? Materialists are determinists, aren’t they?

[Rabbi Michael Abraham] No. A materialist can be non-deterministic. There can be a materialist who thinks there are degrees of freedom in our biological system. He can say that’s possible.

[Speaker D] Not randomness — freedom, really? But freedom is not…

[Rabbi Michael Abraham] Freedom. There are people who claim that a person has free will — rare, but there are such views — that a person has free

[Speaker D] will even though they are not dualists, they are materialists. But 95 percent of materialists, according to their own view, there is no intelligence — is that what you’re saying?

[Rabbi Michael Abraham] According to their method, there is no intelligence. They are mistaken about that, because they do have intelligence. But according to their own position, by my definition, there is no intelligence. That is exactly the point. In other words, this whole world of artificial intelligence and defining intelligence according to the ability to solve problems is a world whose basis really starts from the philosophical question of determinism. Or from absurd materialism, as I discussed — where you don’t really think, you think everything is just currents in the brain. Then all that remains for you to define as intelligence is computational ability — how to solve problems. Then machines have intelligence and animals have intelligence. But according to my view, the concept of intelligence requires two more conditions: exercising judgment and consciousness.

[Speaker D] I see a cat in the street going to catch a pigeon, and suddenly someone else comes from there, and I see it become cautious, and I see it stop, and then it jumps and runs fast. From my point of view it’s thinking, and I think every person would say it’s thinking here.

[Rabbi Michael Abraham] So I say: whether it thinks or not is an open question. None of us knows how to answer it. But what I am claiming is that even if you see the cat behaving that way, you cannot infer from that that it is thinking.

[Speaker D] No, I’m sure it has no free

[Rabbi Michael Abraham] choice. From my point of view, not sure — why? It may be programmed for that, programmed even for that complex process.

[Speaker D] But the process of thought includes for me mental events, and maybe that’s your definition?

[Rabbi Michael Abraham] The process of calculation, not thought. The process of calculation.

[Speaker D] No, I’m saying thought as a whole, as I understand it, includes both the calculation and the mental events and everything together, and that could exist in a cat.

[Rabbi Michael Abraham] Provided it is not deterministic.

[Speaker D] I don’t know why it has to be not deterministic, I don’t…

[Rabbi Michael Abraham] If it’s deterministic, then the one who built that machine is the intelligent one, not the machine.

[Speaker D] It could be that he’s intelligent too, and now the cat wants to catch the pigeon,

[Rabbi Michael Abraham] it’s doing the completion, it feels,

[Speaker D] it experiences all the fear, the hesitations.

[Rabbi Michael Abraham] No, there are no hesitations. That is exactly the point. There are no real hesitations. Even the hesitations are the result of a mechanism.

[Speaker D] True, but it experiences them.

[Rabbi Michael Abraham] Then there are no hesitations. Why do I care

[Speaker D] what it experiences? Those are just illusions.

[Rabbi Michael Abraham] There are no real hesitations, there is no real judgment here. It is all a mechanical process — even the process of thinking this way, weighing that, rejecting this, choosing that — all mechanical. An illusion. That is not judgment.

[Speaker I] The Rabbi…

[Speaker D] argues that the hesitations are what activate this whole part of the mechanical chain — when I feel hesitation, it activates something like that. Sure, but that’s not judgment. True, it activates, but it’s not judgment.

[Rabbi Michael Abraham] The Rabbi claims that anyone who is a determinist has no concept of intelligence — is that what the Rabbi claims?

[Speaker I]

[Rabbi Michael Abraham] His concept of intelligence is a concept of calculation, not of thought. And that is exactly how we get to defining intelligence for water and for a bird and for a machine — yes, exactly that way.

[Speaker D] It empties

[Rabbi Michael Abraham] the concept of intelligence of its essential content and leaves it only as the concept of calculation. And I claim that intelligence is thought, not calculation.

[Speaker I] Intelligence was defined, or the word intelligence, where intelligence is realized.

[Rabbi Michael Abraham] The word

[Speaker I] intelligence originally emerged when human beings used it for creatures that use calculation to solve problems and achieve goals. It’s very simple. I’m saying this without checking, because it’s obvious that that’s the way it is.

[Rabbi Michael Abraham] You think that, and I simply think the opposite. On the contrary, I think the modern definition of intelligence is the mechanical one, because in the modern world we suddenly became materialists and no longer believe in what sets man apart, and then the whole concept of intelligence that accompanied us all along suddenly got thinned down into a collection of mechanical actions. Quite the opposite. I don’t want to argue about the history of the meanings of words — that doesn’t interest me. So call it something else. I am only claiming that the human trait I attribute to human beings as intelligence does not exist in water and does not exist in cats. That is the claim. Call it intelligence or don’t call it intelligence — it’s just a word. You can choose whatever word you want. We can argue about the meaning of words. But I claim that this is not what happens in human beings. That’s all.

I’ll try to clarify this from another angle next time, because this really is a subtle point and it stands at the root of the whole move, so it’s important to me that here we get onto the same page — or at least that I do what I can to persuade you, whether I succeed or not.

[Speaker I] Will the Rabbi get to the stage where he explains to us what consciousness is? What it means to call something consciousness?

[Rabbi Michael Abraham] Solve for us the mind-body problem

[Speaker I] that has been bothering humanity for several thousand years?

[Rabbi Michael Abraham] No, those are two completely different things. The mind-body problem and what consciousness is are two completely different things. But the question of what consciousness is — the answer is very simple, and it’s written in my last two columns.

[Speaker D] Rabbi, can I ask a different question? A question in connection with what the Rabbi said about free choice — something specific the Rabbi spoke about. The Rabbi once said that intuition is what causes the Rabbi, or someone, to believe various things, like free choice and maybe some other things, I don’t remember the examples the Rabbi gave. My question is: after that Libet person proved that in some cases our intuition — we feel we have free choice, and that’s actually not free choice but picking and not choosing, or whatever the Rabbi said — then I already know my intuition failed in part of this story.

[Rabbi Michael Abraham] Why continue believing it in the rest? That too I discuss in the book. First of all, that’s part of the claim. Part of the claim says that it is not true that we feel we have choice. We ourselves feel that this is picking and not choosing. Nobody says: wait, I deliberated and decided to press the button now. I pressed now because now I felt like it. Nobody sees that as a process of deliberation. When a person goes through a process of deliberation and really feels that he is choosing, then the experiments of Schultze-Kraft and Haynes showed that there the readiness potential in fact appears after and not before.

[Speaker E] Okay. Rabbi, given that when we speak about intelligence and artificial intelligence and comparisons between one and the other, where do you place the human ability to use humor or analogies or all kinds of things? That’s part of being human, which artificial intelligence just isn’t in the story on.

[Rabbi Michael Abraham] In my opinion artificial intelligence can tell jokes too, because it imitates human beings in that as well. And if it can’t now, then it will in another year.

[Speaker E] Yes, but to tell artificial intelligence — and I already tried several times — there is no understanding. It asks such questions, like ChatGPT asks questions as if it thinks it’s a question, meaning it doesn’t understand.

[Rabbi Michael Abraham] So I’m saying: then if it doesn’t exist now, it will in another year. That’s not a technical issue. No problem at all — it will be able to tell jokes, it will be able to laugh at jokes if you want, it will be able to do everything. I think it’s already possible now, but even if not, that’s just a technical matter, it will happen.

[Speaker E] So it’s not connected to the mentality of it?

[Rabbi Michael Abraham] No. Only in its case it will happen mechanically, whereas in our case it happens accompanied by mental processes. It simply imitates us.

[Speaker D] Rabbi, I think in this series maybe we’ll need to divide the different parts that we have within thought. For example, just as an example — is remembering, is face recognition, the fact that I know and identify the Rabbi — is that called thought or not called thought?

[Rabbi Michael Abraham] Only if the remembering is initiated. Meaning when I see the

[Speaker D] Rabbi, something that comes up…

[Rabbi Michael Abraham] into my head — that’s not thought. Something that comes to my head by itself is calculation, not thought.

[Speaker D] Even though it is accompanied by the mental?

[Rabbi Michael Abraham] Yes, correct.

[Speaker I] The mental will come.

[Rabbi Michael Abraham] I often wondered, you know, there was a legendary soccer player named Franz Beckenbauer. He was a German soccer star and a libero; he invented that role in soccer. He was known as some kind of genius, an architect of the game. I don’t understand enough to judge it, but in basketball, say, I thought many times about LeBron James. LeBron James really has insane talent. It can’t just be physicality. It’s game intelligence. The way he passes — and I once saw him in an interview after a game, they ask him and he describes everything that happened in the game, step by step, just unbelievable. While he was playing, he was documenting all of it in his head, and he understood what was right to do and what was not right to do, and therefore decided to do this and not that. The interview was hair-raising. Now I asked myself whether that indicates high intelligence. Usually people tend to think yes. I’m not one hundred percent sure, because it may be in his blood. It may simply happen instinctively. Of course it is some extraordinary talent, that is obvious. But the question whether he is exercising intelligence there in the sense I defined here — I am not sure. With Beckenbauer it seemed more like that, because with Beckenbauer he really planned what to do and carried out his plans. At least that was the image — again, I didn’t see enough and don’t understand enough. But with LeBron James I’m less sure. There it may simply be that he has some talent flowing in his blood, and he does it instinctively. I’m not sure that he is — I don’t know, maybe yes — but I’m not sure he’s making some judgment and deciding what to do and what not to do, because everything is happening very, very fast, and then carrying out what he did. If so, then that is intelligence as I defined it. But if it happens instinctively, then I really would not define it as intelligence. Then it’s like bird navigation or the flow of water.

[Speaker I] Does the Rabbi distinguish between instinctive and intuitive?

[Rabbi Michael Abraham] No, I don’t know whether I have diagnostic tools to know whether this is happening or that is happening. I’m speaking hypothetically. If I knew everything that was going on in him and understood that it all happened instinctively, I would not call that intelligence.

[Speaker J] Rabbi, a pilot who operates infinitely many instruments at the same time, divided attention — is that intelligence or automatic?

[Rabbi Michael Abraham] I don’t know. If he exercises conscious judgment and makes decisions, only very, very fast, then of course that is intelligence. But if it is really some kind of natural talent that flows through his veins and simply flows out of him, then I would not call that intelligence.

[Speaker J] But when they test flight aptitude and all those things…

[Rabbi Michael Abraham] That’s not relevant, that’s not… Who said you need to test intelligence? You need to test abilities. Even instinctive abilities — on the contrary, instinctive abilities are much faster, much more efficient, than someone who has to think all the time about what he’s doing; he will usually react more slowly. That’s Kahneman’s System 1 and System 2.

[Speaker C] A few weeks ago we talked about… the Rabbi said that it doesn’t make sense to define love as electrical currents. The question is about animals — animal suffering, say — how do we define their suffering? What does

[Rabbi Michael Abraham] that mean? Their suffering is a sensation of pain, what…

[Speaker C] Yes, but the Rabbi explained that there is no priority to materiality over the assumption of the existence of a soul. Correct, but with animals there’s no barrier to saying there are mental processes.

[Rabbi Michael Abraham] Who said animals don’t have mental processes?

[Speaker C] No, where do they occur without a soul?

[Rabbi Michael Abraham] What do you mean where do they occur? They have some sort of soul, and that’s where it happens.

[Speaker C] No, but the Rabbi always assumes that animals have no soul.

[Rabbi Michael Abraham] No, absolutely not. Soul too — I have no idea, I don’t know. But if they have mental processes, then just as with a human being I think there is something besides matter, with them too there is.

[Speaker C] No, I saw that the Rabbi wrote that there is no conceptual problem with killing animals because they have no soul.

[Rabbi Michael Abraham] No, that’s something else. I claim they have no choice and perhaps no consciousness, although that I don’t know — maybe they do have some level of consciousness — and they have no choice. So in that sense they are not entities with moral standing. But that doesn’t mean they have no mental processes or that they don’t feel pain. If they feel pain, I am forbidden to cause them pain. I am forbidden to cause them pain, but not because they have rights; rather, because I am forbidden to cause them pain.

[Speaker C] Yes, but the very fact that there exists in them some entity that is not physiological — isn’t that enough reason?

[Rabbi Michael Abraham] I don’t think that’s enough reason. So what, they have a non-physiological entity — so what? And this doesn’t pass… I can’t prove such things. It’s a matter of intuition. Someone can come and say he disagrees — fine, okay.

[Speaker C] No, the moment there is an entity that is non-physiological and its appearance here depends on the physiology of that body, then automatically you need some justification for what is done to that entity.

[Rabbi Michael Abraham] Right — but why do I care? Why are spiritual things more significant than physical things? This exists and that exists — what’s the difference? The question is where you place the focus in what sets man apart. I place the focus in what sets man apart on judgment and choice. That is what sets man apart.

[Speaker C] And why is that the point that grants moral rights?

[Rabbi Michael Abraham] Good question. I don’t know how to answer it. It’s a sort of basic intuition. Fine — someone may come and say he

[Speaker G] doesn’t agree, not…

[Rabbi Michael Abraham] okay, fine, I don’t know how to justify such a thing.

[Speaker G] There is

[Speaker E] some kind of

[Speaker G] desire, it’s

[Speaker E] not from here.

[Rabbi Michael Abraham] Okay, wait.

[Speaker E] Wait, I just don’t want

[Speaker G] to give up on any of you. Rabbi, can you hear me? Yes. Yes, I want

[Speaker E] to ask a question that is somewhat connected to Jewish law from what we discussed. There are all kinds of living creatures like bees, aphids, ants, that are produced like a computer from… it’s called cyclical parthenogenesis. That is, they are not produced by a father and mother, but by some entity like the queen bee or something like that, without sexual reproduction.

[Rabbi Michael Abraham] Okay, nice. Very nice.

[Speaker E] Now on the Sabbath, for example, when there are ants — and certainly only the workers are there — they are supposedly like a computer. They’re in the sink or in the path where a person walks or in all kinds of places where, on the one hand, if I open water and it flows, that is taking a life, but on the other hand there’s nothing there, meaning they weren’t born from male and female.

[Rabbi Michael Abraham] The fact that they were not born from male and female does not necessarily determine their status. The Talmud says that a louse, since it is not produced through sexual reproduction, there is indeed no prohibition against killing it on the Sabbath. But on the moral level, or even on the halakhic level, I’m not sure — who says that is the criterion?

[Speaker E] I don’t know, I’m relating to it — it’s simply a question that occupies me. Now it’s summer and everything comes out, right? And for example the ants.

[Rabbi Michael Abraham] If you do it unintentionally or as a mere preoccupation, then fine — then you did not violate a prohibition because it was unintentional, not because it isn’t taking a life. Do you want to argue that even intentionally it isn’t taking a life?

[Speaker E] Exactly, yes — that’s the issue I’m asking about. What do you think?

[Rabbi Michael Abraham] I don’t think so. It seems to me that it is taking a life as well. It’s a living creature; it doesn’t matter how it came into being.

[Speaker E] That’s a question. The Sages saw this as a kind of process. As if if it came into being like that without father and mother, without sexual reproduction, then…

[Rabbi Michael Abraham] The question is what the product is, not how it was produced. I don’t know — if you combine sperm and egg but do it in a test tube, entirely without a woman’s womb, all in a laboratory — is the human being that comes out of that so clearly not a human being that you’d be allowed to murder him?

[Speaker K] No, absolutely not, but I’m asking where, between the Sages and this, where is the weight?

[Rabbi Michael Abraham] So I’m saying: I think the question is what the result is, not how it was produced. And if the result is an animal, then I think you may not kill it. By the way…

[Speaker D] I think with ants it is sexual, I think there are males there that fertilize the queen.

[Speaker E] No, absolutely not, it’s complete parthenogenesis — meaning from itself, that’s it.

[Rabbi Michael Abraham] Fine, okay, we can check that afterward, but on the principled level I don’t think the mode of coming into being matters here. Okay. Good. Okay, fine. Goodbye, Sabbath peace.