Outsourcing Humanity: Ancient Insight for the Age of AI

Transcript begins at 8:56 in the video.

AI: A Difference of Degree or of Kind?

Geoff Mitelman: Thank you, Ora. Thank you all for joining here this afternoon, this evening, wherever you are. It’s a joy to be with you all, and I’ve loved AJR for many, many years.

As Ora had mentioned, my work is in the world of science and religion. and I started an organization called Sinai and Synapses a little more than 10 years ago, looking at a variety of different questions surrounding Judaism and science, and I would say about 75% of my work, 80% of my work right now, is thinking about Judaism and AI, and thinking about this question that we’re looking at today: What does it mean to be human? Where does AI come in and look at questions of existential issues of “who are we as human beings” and also practical questions: What are the ways in which new LLMs and agents – what they call “agentic AI” – how is that changing who we are?

And so what I want to look at, and we can start with, one of the slides here. My favorite definition of technology as a whole is technology is whatever got developed after you were a kid. And so for me, I’m in my late – I guess late – 40s now at this point. So, when I was growing up, a lot of the big fear was the rise of violent video games. But depending on what age you were, it may have been smartphones, or the internet, or, depending on how old you are, television.

But one of the ones that I love to talk about is The War of the Worlds. The War of the Worlds, which you may know, was a novel written by H.G. Wells in the late 19th century. And Orson Welles, in 1938, created a radio play of the War of the Worlds – told that story there.

And what we think about now in 2026, the story that’s often told is that the Orson Wells broadcast of the War of the Worlds was that there was mass panic in the streets. As we see, there was an article in the New York Times, it said “radio listeners in panic, taking war drama as fact.” The truth is, yes, there were some concerns and people were calling the radio stations and they were calling police there. There was a little bit of concern that was happening of what was going on. People didn’t really know what was happening. But there was absolutely no mass panic in the streets. That was a story that was fabricated. It was made up in a lot of ways from the newspapers. The newspapers were really trying to be able to say, “Don’t get your news from this newfangled invention called the radio. You should get your information only from us, from the newspapers here.”

And we can think about how that changed the media landscape, where, for example, radio – you bought a radio one time and you never had to buy it again, or maybe once every 10 or 15 years. The newspaper you had to buy every day. The radio gave you instant information. The newspaper was a little bit later, and so gave a little time of reflection. And what the new all these newspapers, including the New York Times, what the real message was that “new technology is scary, don’t use what comes into this new world, use what we’ve done here.”

And I think that’s a lot of the questions that we are grappling with as thinking about AI. How is AI changing who we are? And a question that that we’re grappling with – and I’ll come to this towards the end, of is the question of “Is AI as a technology a difference of degree, or is it a difference of kind?”

We’ve gotten used to all sorts of new technologies here – again, whatever you were used to growing up, that’s just part of your daily life. And we will use the word “technology” or “AI” as kind of a boogeyman, as something that is scary, something we should not use, and yet it very quickly it becomes integrated into who we are. The challenge is what happens as the new technology gets adopted more quickly and more widely. So if we can go to the next slide there, of a text that we come from, Bereshit, from Genesis 2, where God has created Adam. And it says – I’ll read in Hebrew here as well – that it says that “God, Adonai, settled the human in the Garden of Eden to work it and to tend it,” l’ovdah u’l’shomrah.

And Rabbi Joseph Soloveitchik also talks about this – that lots of Jewish thinkers have thought about this, that we are commanded, ovdah, to work, to be able to be technological. That comes in, that we are designed – if we think of what technology is. A another good definition of technology is “the human ability to take the natural world and make it better,” make things cheaper, make things easier. We are commanded to work the land.

And that is part of our evolutionary history as well. One of our human cousins, the Homo habilis, is the human being that was able to be the handyman. So we work, and we create the world, and yet we also have to l’shomrah, we have to protect against – well, a lot of rabbis think we’ve got to protect it against the ovdah. We have to work it, we have to protect against the potentiality of overwork that comes in, because, even though we are told to work and to guard the land, the Garden of Eden – “God commanded the human and said, ‘Of every tree of the Garden of Eden, you are free to eat. But as for the tree of knowledge of good and evil, go good and bad, you must not eat of it. For as soon as you eat of it, you shall die.’”

This was a little revelatory to me when I first heard this thought, and when you hear it, it’s obvious in retrospect – which is that technology always outpaces ethics. I’ll repeat that: Technology always outpaces ethics. We don’t know if something is going to be good or bad for humanity until it exists out in this world. Adam is commanded to work the land, and only then will he find out “Is what I’m doing going to be good or bad?” We don’t know how the world is going to be until there is something new out in the world.

And yet, one of the things that comes up that we think about of humanity, we are grappling with: What does it mean to be human? Is there something that makes us uniquely human that comes in?

And if we can go to the the next slide, Rashi, quoting from from Chapter 2, Verse 7, where “God formed the human Adam from the soils, adamah, blowing into his nostrils the breath of life, the human became a living being” – a nefesh chaya, is what the Hebrew says. And Rashi talks about how also cattle and beasts are called nefesh chaya. We have something in common with other animals, but the nefesh of humanity is the most highly developed of all of them, because the humans were granted understanding and speech.

We tend to think that what makes humans at least unique or special is our ability to have language and speech. The question of AI, the word “AI” is actually a word that is that was coined at Dartmouth in the 1950s. People have talked about artificial intelligence in a variety of different ways. If you’ve used Waze, you’ve used artificial intelligence, because it’s able to guide you through traffic. If you’ve used Netflix 10 or 15 years ago, looking at an algorithm, you used AI. If you were looking on your social media, there was AI being used there to push what comes to the top of your feed and what doesn’t.

But over the last three to four years, we’ve conflated the words of AI and a lot of what we talk about as LLMs – large language models like ChatGPT and Claude – because we tend to think that the element of understanding and speech is what makes us human. The interaction with a prompt – with something written – makes it feel like it’s human. And that’s where it’s starting to feel a little bit scary.

What’s also interesting is in the 1960s, there was a a computer called ELIZA that was programmed as what’s called a Rogerian therapist. Rogerian therapy basically just says, “Oh, interesting. Oh, tell me more. How does that make you feel?” There were maybe six or seven prompts, maybe more – maybe 10 or 15, but not many prompts. And people were sitting in front of their computer typing back and forth, thinking that they were getting therapy from a chatbot. This was in the 1960s.

The explosion of what LLM are able to train human thinking on – human writing – that’s what makes us feel like, “Oh my god, I feel connected here.” The language and the speech are aspects of what we think makes us uniquely human, and that’s why we’re feeling worried about “Is AI going to replace us?”. Questions of creativity – we’re not worried that a dishwasher is going to replace us, but we are worried about what may happen as language and speech come in.

So just as the last slide, as I’ll conclude here, as I notice AI is a difference of degree or is it a difference of kind? Because we get used to new technologies, but the speed of adoption is unprecedented in terms of how quickly Americans have adopted chatbots. We treat technologies as tools, but LLMs feel relational and human. We’re used to technology solving a specific and narrow problem, but some private companies are trying to develop what’s called general intelligence, solving any kind of problem, or even super-intelligence. And we can talk about that later.

And then, lastly, we’re used to developing tools that meet and goals our needs, but AI’s goals may not match with humanity’s goals. These are questions that are open questions. These are ones where Judaism brings wisdom to this conversation. And I’m excited to hear from Andy to expand on these conversations as well.

AI and the Classical Idea of the Person

Ora Horn Prouser: Geoff, thank you so very much. That was really a great beginning for this conversation, opening up a lot of very important language and tools that we need. And I pass it on now to Andy to take us to the next steps.

Andy German: Thank you very, very much. And I apologize if you hear an air raid siren – uh, it’s here in Israel, so I’ll try to make sure it doesn’t happen. Here’s what I’m not going to talk about in the few minutes that are allotted to me. I’m not going to try to answer the question of whether AI models are conscious or intelligent, or the question of “Do we have ethical duties to them?”, or the question of the effects of AI on higher education or the economy. I won’t do that, because some of the answers to those questions seem obvious to me. For example, are AIs conscious? No. Might they become conscious? No. It’s a categorical mistake.

But the second reason I won’t discuss those questions is that, in order to think about them and in order to think about AI, we need to answer two much, much more fundamental questions that aren’t about AI at all. They’re about us. And those two questions are: What exactly is a human being? And: Why is it even good that human beings exist? What I mean to say is that AI is a remarkable technology because it forces us to philosophize, whether we like it or not. And I imagine there may be people who don’t, but I am a philosopher, so it’s good for my business.

Here are two answers to the question of “What exactly is a human being?” Elon Musk human beings he says are biological boot loaders. He means by that that human beings are a tool for it’s like a disk on key, a tool for kick-starting artificial intelligence by uploading the fruits of human cognition into AI models. Elsewhere he calls us a source of willpower and intentions kind of raw material for use by someone or something.

Marvin Minsky, who was the founder, really, of the field of machine learning, called human beings “computers made of meat.” Now, you could say that those answers are striking or ridiculous or even downright terrifying. But I can assure you that they are 100% unintelligible unless you understand two things. One, those answers would have been completely inconceivable prior to about the year 1600. And two, they result not from some technological breakthrough, but primarily from a profound philosophical transformation. And to think about AI, we need to know something about that colossal transformation.

So here goes 2,500 years of philosophy in about eight minutes. This should be no problem at all. Let’s start with a very, very old, a very dead, very Greek man, Aristotle. Aristotle wrote the first ever systematic treatise on ethics. And in that book, he tries to understand “What is the good that we want to achieve by all of the choices that we make in life?”

Now, by name Aristotle says we all agree on what the good is that we want to achieve. It’s happiness. That’s what we want. We ultimately want to be happy. For each choice that you make, I could ask, “Well why did you choose X?” And you’ll say, “Well I chose X to achieve Y.” And then I’ll say, “Okay, well, why did you choose to achieve Y?” And you’ll say, “Well, because of Z.”

And we could go up the chain of reasons, and eventually we’ll get to your answer is “Because I want all these things, because I want to be happy.” And then I don’t need to ask, “But why do you want to be happy?” That would be a stupid question. Now, this is all well and good, but happiness is just a name, and people don’t agree about its content. There are 9 billion people in the world, so it seems like there might be 9 billion definitions of happiness. It would be an impossible task to know what it really is.

But Aristotle disagrees – and if we can cue that first slide, we’ll see why. This is a piece of text from Aristotle’s Nicomachean Ethics. To define happiness, Aristotle says we need to first identify – or, to define human happiness, we need to identify what is human about us. And watch how Aristotle does it. He says: “Humans are beings. What is a being?”

The answer to that has at least two parts. To be anything at all, for Aristotle, is to be something that has a distinct identity. And especially for living beings, to be is not just to take up space. It’s to be busy doing something. Each kind of living being is defined by some distinct activity. We do a lot of things. We’re active in a lot of ways. We eat, we move, we perceive. Which one of these activities, asks Aristotle, picks out what is human about us? Just as we identify, you know, what a flute player is by what he does, and what a sculptor is by what she does – the same must hold for the human being, since, as far as I know, you have to be a human being before you can be a flute player or sculptor.

So, to define the specifically human happiness, Aristotle thinks we need to find the work, the activity, which only we humans do fully, and which, when we do it well, we’re most fully ourselves. And as you can see from the text here, Aristotle thinks that the unique work that makes us human is what he calls “the activity of the part of our soul that has logos,” which is the Greek word that I translate [as] “articulate speech.”

What does Aristotle mean by that? Of course, there’s no articulate speech if we don’t have rationality. And that’s why you often hear that Aristotle defined the human being as “the rational animal.” But you should be careful. Aristotle doesn’t just mean that we are rational because we can speak, or build fancy theories, or know a lot of stuff. He means a lot more than that. Being rational shapes us completely. We have wonderfully flexible hands with opposable thumbs because we’re rational, not the other way around. We walk upright because we’re rational – it’s not that we happen to become rational because we evolved to walk upright. We’re rational when we learn; we’re rational when we produce things according to rules – the capacity the Greeks called techne, from which we get our word “technology.” We’re rational when we act justly or courageously. We’re rational when we watch Sophocles’s Oedipus Rex.

I want to note four things about what Aristotle says here. One, you cannot be human if you’re not actively busy being human. Humanity is not a collection of properties that just happen to be lying around in the same sack of flesh and bones – it’s a unified activity, humanity is, an activity with multiple expressions.

Third, the activity is dictated to us by nature. It’s given. It’s not something that we choose or create.

And fourth, because being actively rational is our humanity, “when we’re rational in an excellent and beautiful way,” as he says at the end of that text, “we’re most fully human and therefore most happy.” And Aristotle even says a little bit more than this elsewhere. “The world,” he says, “is understandable. It has an intelligible structure. When we exercise our reason, we understand it, and therefore in some sense complete the world, make the world fully what it’s meant to be, which is why it is good for human beings to exist.”

So for Aristotle we, as you can see, we aren’t a bootloader mechanism or a computer made of meat. How did then we get from Aristotle to views like those of Elon Musk and Marvin Minsky? Obviously, there were changes in religion and politics and technology and culture. But underlying all of those changes, there was a change in how philosophers thought about nature and reason and humanity. Nature ceased to be the encompassing given of which we’re a part. It became an object, and we stand over that object as subjects. reason became primarily productive or utilitarian.

For Aristotle, the main goal of understanding is to understand. For someone like Thomas Hobbes, for example, the main goal of understanding is power. So, let’s take a a look at a very short quote from maybe the greatest of all philosophers who contributed to these changes and reflected on them. And that would be René Descartes, the “I think, therefore I am” guy. If we could have the second slide. There he is – a very – you wouldn’t buy a used car from this fellow.

The text that I have here comes from Descartes’ Discourse on the Method, which is a book in which Descartes describes to us, very sort of revealingly, how he became Descartes. And not what he says there – he says, you know, in his studies, he was mesmerized by the simplicity and the absolute necessity and clarity of mathematical and geometric proofs. Mathematics was his model and standard for knowledge. What you cannot know like you know mathematics you don’t really know. Descartes is also concerned, as you can see, with new kinds of principles of knowledge – knowledge relating to physical things, not speculative arguments. And Descartes realizes that if we combine this mathematized model of knowledge with physical and practical goals, we could make ourselves what he calls there “masters and possessors of nature.”

It’s an amazing phrase. Why should we want to master and possess nature? Well, Descartes answers. He says, “In order to use nature, to make an infinity of devices, to cure an infinity of illnesses, to make our lives infinitely comfortable.” And make no mistake, if you look at the end of that text, one of the things that we will master is our own nature. We’re going to remake it so radically, he says, that we may even conquer old age.

Descartes was one of the first great thinkers to redefine philosophy as mathematical natural science in the service of power. Aristotle knew about techne, and he knew about mathematics, but he never thought of philosophy as the combination of techne and mathematics. And without that redefinition, there would be no AI. Now, note carefully: The intention to dominate nature preceded the technologies that gave us the power to do so. I hope you all see that.

Now, it’s clear that Descartes triumphed bigly, right? Since the domestication of electricity, basically we’ve conquered nature, and we live almost entirely in an artificial world. We live in Descartes’ world. We’ve turned that awesome scientific power onto ourselves, and we look at our humanity not as a unity given to us by nature or by nature’s god, but as a collection of properties that we could separate, tinker with, reform, and discard at will. For a while now, we’ve done that with bodily organs. And now we’ve created this eerie copy of some of our cognitive powers. We’ve made something that looks incredibly like it’s learning and reasoning and planning and speaking – even though it has no awareness of self or time, it has no relation to a body, it has no desires, it has no fears, it has no questions. And this amazing copy, when you marry it to immense computing powers, is on the verge of making certain kinds of cognitive work, certain kinds of human knowledge, essentially free, which means they will have no value in the market.

And when you combine that power with advanced robotics, which is advancing every day, it may make almost all of our human work superfluous, because something else will be able to do that work better, or at least well enough, and without needing a salary or without getting hungry or bored or sick or worried – you know, and having a midlife crisis.

Now, what should we think about this reality that is being created? Some people promise us that it’s going to be like living in, you know, the big rock candy mountain, where the rivers all flow beer and there’s eternal health and sunshine. Other people tell us AI will be smarter than all of us and make us obsolete. And still other people tell us it’s not a big deal. It’s just another tool.

Basically, what we need to decide is which of these is true, and whether we want any of this, and how much of it we do want. But here’s the catch: We can only do that if we ask, again, what it is that we have to be doing in order to remain human, and what is inessential and could be outsourced?

So it looks like we spent 500 years thinking that we were moving away from Aristotle, when in fact we were going back to the same questions that he had, which is – history of philosophy is funny that way.

By the way, I’ll save you the trouble. there’s literally no point in plugging these questions into Claude or ChatGPT. It’ll spit out answers, but you’re going to have to judge if the answers are any good. And it’s only possible if you already know what it is to be busy being human. So, you’re just going to have to do this one yourself. Thank you very much.

The Myth of Conscious AI

Ora Horn Prouser: Thank you so very much. Really, you have us all thinking right now, which I hope is the goal here. So I want to open first, and first of all, if people in our audience have questions, please feel free to put them in the Q&A function and we will get to as many of them as we can.

But I want to start with a few questions to both of our panelists – and I think I’m going to start by following up with where you just were, Andy, with the question about – that you really don’t think we should speak of AI as intelligent or conscious, and that that’s a categorical mistake. So, could you just say more about why you think AI is neither intelligent nor conscious, and why this should matter? And then, of course, Geoff, please respond after Andy as well.

Andy German: Sure. To say that AI is conscious or could become conscious or intelligent, I think, is to make several hidden assumptions that are so hidden to modern people like us that we don’t even know they’re there. It’s to assume, for example, that consciousness and intelligence are things that are unconnected to having an organic body, that consciousness is something that develops or could emerge from non-conscious parts, and that intelligence equals possessing a great deal of knowledge or data. Now, all of those are assumptions that are unexamined, and I’m about as convinced as you can possibly be that they are not even wrong. They are too incoherent to even be wrong. The fact that we have those assumptions is a product of just a catastrophically confused way of thinking about what people call the mind-body problem.

So, I don’t think AIs are conscious. But I actually don’t think – and I’m not going to bore you with the philosophical arguments – you know, it bores other philosophers, so I wouldn’t put you guys through it. But in a sense, it really doesn’t matter if AIs are conscious or not. Because what’s more dangerous, I think, is for people to act as if they’re conscious. For example, you now read about people who call their AI models their “best friend,” you know, their therapists. And there are companies now dedicated to making human-sounding and human-looking robotic romantic partners.

Some people talk about AI as becoming so smart that they’re going to relate to it as though it were a divine being. Now, what does that tell you? It doesn’t tell you anything about the AIs. They’re not conscious, and they’re not divine. But what it tells you is that there are a lot of human beings out there who are so desperately looking for ways to offload their humanity to someone or to something else. There are a lot of people, I think, who find their humanity to be a burden, and that’s the danger. Not that AIs will ever become conscious, robotic monsters, but that we won’t be able to tell the difference, I think.

Ora Horn Prouser: Thank you. That really is terrifying. Geoff, do you want to expand or respond at all?

Consciousness and Intelligence

Geoff Mitelman: Yes. And I want to build off of that. So, I tend to agree with you on on consciousness, although I would push back a little bit on the question of intelligence, because one of the things that you also bring up is that we often conflate intelligence and consciousness, and they’re not necessarily the same thing. And if you define intelligence as “trying to achieve a goal of some kind,” I think there there is an aspect where you could see AI has a form of quote-unquote “intelligence.”

But you bring up a really important point, Andy, which is the way in which it’s changing our humanity. And one of the things I’ve been thinking about a lot is how we talk about Martin Buber. And Martin Buber talks about humans as not just when you say “I.” There’s an “I-it,” and there’s an “I-thou.” And we tend to actually view most human connections, understandably, as an “I-it” relationship, where it’s a transactional aspect of it. And that’s not a bad thing. I have to get my dry cleaning done, and I treat my dry cleaners [as I-it]. I like them, but I don’t have a deep relationship with them like I do with my spouse and my kids.

And I think we, for most of human history, understandably, have had an I-it relationship with technology, kind of by definition. We use technology to meet something that we need. And I’m not as worried when we are using an LLM in an “I-it” relationship. I think there are questions that come up about writing and intellectual property and education. Those are real questions there. The danger in my mind is what happens when we are viewing an LLM as an “I-thou” relationship, because there is no “thou” coming back to you.

Andy German: That’s right.

Geoff Mitelman: And it certainly feels that way – that it’s not a functional question of, “Does this appear to be human or not?” There is value in having friction in our lives, and one of the areas where we want to make things a little bit difficult, of interacting with human beings that are that are messy – because that’s what human beings are. And my hope, by the way, is that I think actually there may be a reversion back, where people are going to recognize how important and valuable the human interaction is. I think we’re still figuring it out, but we evolved as needing human connection, and I don’t think we’re going to lose that need.

Andy German: I agree.

Learning in a Post-Literate Society

Ora Horn Prouser: Yeah, I hope not. Thank you. Geoff, I think I want to start this question with you. How do we learn when there are new methodologies for learning? And what is Jewish learning going to look like in what some have called a “post-literate society”? And how do we even think about literacy versus orality? So, you can speak to that.

Geoff Mitelman: Thank you. So that raises a variety of different questions. One of the things that if you talk about and look at LLMs, and the question of “Oh my god, does it quote-unquote ‘feel human’?” Most of that is going to [be felt] using an LLM mediated through a chatbot, and that’s because the writing feels human. We’re not – and I don’t think we’re ever – going to be able to feel as connected with an oral conversation, because we can go off on tangents, we build off of discussions that come in. Conversations exist in an oral tradition. And so – and we’ve seen this, by the way, if you’ve used an LLM, I do this – if you do a speech-to-text translation, it’s a lot easier to just type it because of the number of times where I’ve gone “Is this correct?” and “How do I spell it?” The orality allows you to be able to fill things in that you may not know.

And that’s how a lot of education has happened for thousands of years. Much of Jewish thought – and Andy, I know you’re much more of an expert on this as well, in Greek thought – it was being able to share information. And there was a discussion in Greek thinking, “Should we write down what Socrates had to say?” In Judaism, we have the written Torah and the oral Torah, but the oral Torah got written down as well. So what happens when all of our written corpus comes, and what is our job as educators sharing not just the information, but the relationship that comes in? I think that’s one of the things we’re teaching.

And language – it is not just designed to be able to communicate data and information; language is designed to be able to build relationships. And so that’s where I think the question that comes in, of relationships with chatbots – some of that is, “How do we find time with people in the real world, in physical bodies, of seeing people’s faces, people’s body language, the tone of voice?” That is so core to our humanity that, I think, that’s an area where, again, we will be able to claim an element of humanity, of whether or not it does or it doesn’t – it’s an area where we can say, “This is something that we want to claim as keeping human.”

Ora Horn Prouser: Thank you. Andy, would you like to expand or respond?

Andy German: I’ll be very brief. I’ll just say one thing. I think what we have to worry about, primarily, when it comes to learning, is the difference between those of us who grew up before LLMs were integrated into everyone’s phone and computer screen, and those of us who will grow up without having known a world that does not have that kind of integration. And the difference is this: Rabbi Mitelman and I are of a generation that have grown up learning how to speak and read and think and write, and then can go and judge how the AIs do these things, and whether we want to use them for any of those technologies, and to what extent we’d rather maintain a human connection, which I think human nature definitely pushes towards.

But the generation coming up, for example, into universities – we were warned just the other day – in Israel, people come at a later age to university, but we’re soon to get students who were born after the iPhone. And I don’t think we yet know the degree to which these technologies will atrophy certain skills that we assume are part of human nature. But they are, you know, susceptible [to] being atrophied. We’re going to have to find out just how atrophied they will be, and how much of a struggle it will be to maintain, for example, inter-human oral communication when people have not sort of learned how to be articulate by reading and writing and thinking on their own. It’s going to be interesting, I think.

Ora Horn Prouser: Yeah. If we were to expand on what you both just said, in terms of the gap between the generations, which I guess there’s always a generation gap, but this one seems to be of a different nature. Or maybe they always say that as well, um, but this one seems to be on a different nature. So what do we imagine a Jewish childhood will look like in 10 or 20 years, in terms of learning, in terms of community? How do we imagine that or even can we imagine that? I’ll take either one of you.

Geoff Mitelman: I’m happy to jump in again – you know, thinking of this question of “What the role is of information versus relationship?” And there are aspects of both of those questions. One of the things that we need to recognize is, there’s a difference between learning versus studying. At least in my mind, more than Judaism is about learning, Judaism is about studying. It is [more] about the process of looking at the text and grappling with the text than necessarily knowing the quote-unquote “answer.” It’s the process by which we explore these kinds of questions. And actually, AI can be a wonderful way to expand these questions and discover more information, which then leads to new avenues of looking at questions there that come up.

You know, another example that that we can think about is the difference between searching versus browsing. And I think that’s a skill that we need to be deepening – searching, of when you go on Amazon, of “I need this particular book,” – “Here’s its name, bam, you’re done,” versus walking around a bookstore and you go, “Wow, that’s an interesting looking book I never would have thought of that before.” That’s why, for example, having the structure of Daf Yomi or Parasha HaShavua – like, we don’t get a choice of what we’re studying, necessarily, and we do have a choice of what we want to study in our other times as well. So seeing the connections, that’s going to be an even more important conversation and question.

And there’s a professor that I like named Daniel Willingham, who wrote a great book called Why Don’t Kids Like School? And one of the things that he said – one of the problems with school is that “We give kids the answers, but don’t explain to them why this is a good question.” And I think that’s part of what education is going to need to be: How do we show what is a good question and why it’s relevant?

Ora Horn Prouser: Thank you. Thank you very much. Andy, would you like to add or respond?

Andy German: Well, no, I’ll let you ask the next question. I don’t want to steal the thunder of someone who’s an expert in Jewish children, since although I am Jewish and I have children, I’m not an expert in them.

Looking Inward Toward Humanity

Ora Horn Prouser: So, thank you. I think we’re going to just ask each of you, and then we’re going to open this to the audience questions, but if you can answer this in 30 seconds, which is asking a lot of you: What does this moment ask of us as humans? So, a very short answer to that. Andy, do you want to go first?

Andy German: Sure. I’ll tell a story that I saw in a book by the great physicist Werner Heisenberg. He tells the following story: Imagine a warship that has been built to – you know, an iron warship, that, in order to withstand the enemy’s bombs and shells, they keep adding more iron siding to it to make it impregnable, until the point is reached in which the magnet of the ship’s compass is no longer responding to the polarity of the earth, but to the iron of the ship – in which case, of course, the ship becomes impossible to steer. It’s just going in circles because it’s responding to its own polarity. What is the captain supposed to do in such a situation? And the solution to this terrible problem is simply to realize that he needs to look again at the stars, and use them as his compass. I think what we need to do as human beings is not fear this technology – well, we could fear it a little bit, but primarily we have to keep an eye on our own nature. That’s the star that will tell us what use to make or not make of it.

Ora Horn Prouser: Beautiful. Thank you so much. Geoff, how about you?

Geoff Mitelman: So, here’s a theology – and I don’t think it’s a particularly good theology, but there’s a theology from the history of science called the “God of the gaps,” that “God does whatever science can’t do right now.” And the problem is that science progresses and progresses and progresses, which means that if you don’t have a good, robust theology, God becomes smaller and smaller and smaller.

Andy German: He’s unemployed. (Laughs)

Geoff Mitelman: And there’s an idea that they talk about as “the human of the AI.” We keep saying that humans are whatever AI can’t do right now, and I think there’s going to be more and more of what AI is going to build in. So we need a robust theology and a robust anthropology that can look at these kinds of questions.

And one of my friends and colleagues, Dr. Sam Arbesman, has asked – and I think this is a great framing – is, rather than asking what makes us uniquely human, we should be asking what makes us quintessentially human. What are the things that we can, as our core, of “We want to make sure that this stays as a quintessentially human activity.” Even as AI gets better and better at that, and partners with our human capacities, what are the ways in which we want to stay – what aspects we want to keep as quintessentially human, rather than uniquely human?

Q&A

Ora Horn Prouser: Thank you. Thank you both so much. We have some questions from the audience that I would like to bring to you. The first one is from David Breslauer: “You have discussed LLMs, which use statistics to spit back existing information to some sort of parsed inquiry. What about creative AI systems? Where are we on the development of systems that can take existing knowledge and create new knowledge from it without being directed to do so – that is, where a machine can take existing knowledge and come up with a new idea that does not exist already?” Either one of you?

Andy German: It’s sort of – you know, generative AI, I suppose you could call it, but generative AI on steroids. We’re not there yet, to the degree that I am aware, but I don’t see that in principle it’s impossible to create such a device, because creation, except for God, doesn’t happen out of ex nil, right? Only God can do that. Human beings always create on the basis of the reality of their experience of prior creation, the knowledge that they have.

Now what we’ve basically done is plugged all of those things into effectively unlimited computing power. And, of course, there’s an element here of “There’s someone behind the screen.” We also have human beings behind these programs doing reinforcement learning and pumping in human preferences to make the LLM models know what kind of answers we’re looking for. So I don’t think we’re going to have pure creativity from AIs, but we’ll have something that looks close enough to it that, for many people, and perhaps for many corporations, if you have to choose between a human creator to whom you have to pay a salary and to whom you have to give sick days, and an AI creator that may not be Leonardo da Vinci, but will be close enough for, you know, government work – if you think people are going to choose the human over the machine, I can sell you a few bridges.

Ora Horn Prouser: Thank you. I’m going to keep going with questions. Maybe we’ll do just one at a time for these questions. “What is the rabbinic responsibility to address these issues, precisely in the post-literate world, to ensure that the Jewish people still deserve the mantle of ‘the people of the book’?”

Geoff Mitelman: So, it’s a great question. I’m going to give an intentionally provocative answer, and I’ll expand on that a little bit. My immediate reaction is “none,” because the way in which we communicate that information – there’s a piece in our CCAR Journal on this, about “the medium is the message.” When we study Torah, for example, we study – if you’re prepping for Torah study, you are not getting out a Sefer Torah from the ark and rolling it to the right place. And one of the things that drives me nuts in pop culture, when they have a rabbi who’s studying something, and they’re like, “That’s not how I look at that text.”

The way in which Sefaria has changed the way in which we explore different questions, and the printing press, and the Vilna Shas, have changed the way in which we look at conversations – I don’t think that the book, as a specific medium, is sacred in that way. I think there are questions of “How do we communicate?” And I think that the learning is there, and I think there needs to be a deep conversation and relationship that comes in. But I’m not averse to – I see some wonderful memes of Daf Yomi that give a little insight into what the Daf Yomi is for that day, which not the same as actual studying and deep work that comes in. That’s where I think there is still a place for rabbis – of being able to spend time thinking, reflecting, exploring. But I don’t necessarily think that that books are going to be sacred in the same way that sacrifices are. That was a technology that existed 2,000 years ago and, “Oh my god, what’s going to happen if that goes away?” Well, we created something new. So, that’s an intentionally provocative answer that I’m happy to give more nuance offline, but that’s my immediate reaction.

Ora Horn Prouser: Thank you. Andy, did you want to add to that or should I keep going? Keep going? Okay. “Some of the major concerns about AI include the environmental impact of powering the technology, and the impact on the workforce, as we have seen layoffs in certain companies that shift to using AI. So, what does the Jewish ethical tradition contribute to these concerns and this conversation?” Either one?

Andy German: Rabbi, do you want to go first? Do you want me to speak about the Jewish [philosophy]?

Geoff Mitelman: You go ahead.

Andy German: I would say this: there’s no dignified – you know, Judaism and Greek philosophy don’t agree on many things. But I think there’s a point in which they’re not identical, [but] they come within sort of a speaking distance of one another. There’s no human life, no dignified human life, without work. The kind of work may change, but there’s no dignified human life without work. You know, Aristotle’s point is that we have to be active. There’s a certain work that goes into being human. And it includes, among other things, the work of creating, the work of serving… so, there’s no dignified human life without work.

And therefore, we have reason to fear that if human cognitive work and human labor becomes so cheap as to be effectively superfluous, the incentive to shift out human beings and shift in AI or robots will be overwhelming. But we can resist it, so long as we remember that work is a source of respect for the human being, and we have to strive to build an economy where that fact – or that value, sorry – is a fact to the extent possible.

Ora Horn Prouser: Beautiful. Thank you, Geoff. Did you want to add?

Geoff Mitelman: No, I’ll take that. There’s actually a piece that – I really respect a journalist named Sigal Samuel who writes for Vox. And she wrote a piece – and I was teaching about this in my AJ class yesterday – about what “How do we find meaning in a world where work is a very different aspect?” And I think there’s an element of melachah versus avodah versus amal, of “What does it mean to striving?” What is the production that comes in? And I think there’s there are going to be elements where what we really value is the human imperfection that comes in, and valuing the the creative nature, and the importance of pausing from that creative nature as well.

Ora Horn Prouser: Beautiful. Beautiful. I think that’s probably really a good place for us to bring this together with really understanding the value of human beings, and what we need to maintain our value. And I very much appreciate what you both brought to us. There have been many more questions. Those of you who asked questions that we didn’t get to them, we understand the frustration, but we appreciate your questions nonetheless, and we really appreciate you being here with us.