Striving to be Human in the Age of AI -- Rosh Hashanah Day 1 5787

‘Tis the season for confession and making amends. And in that spirit, I want to say aloud today something that has been weighing on my heart: I owe my friend Abe a long-overdue apology.

Some of you know Abe. He is one of my oldest and closest friends -- more like a big brother and mentor to me than anything else. He is also a rabbi, and many of you met him when he officiated my installation ceremony here in the spring of 2025.

To understand why I need to apologize to Abe, I need to take you back to 1998: Ross and Rachel were on a break. The internet was AOL and announced itself by screaming through your telephone. And the future of home entertainment came on giant silver frisbees called LaserDiscs.

For those of you who don't remember LaserDiscs, imagine a DVD the size of a vinyl record, because for some reason the future used to require a great deal more shelf space. I was the only person in our group of friends -- actually, come to think of it, maybe in the whole state of Georgia -- who had a LaserDisc player, which was hooked up to what was at the time just about the biggest TV you could find, and a very elaborate surround sound system. And for some reason, we had a copy of Blade Runner on LaserDisc -- director Ridley Scott's early-'80s sci-fi classic, starring Harrison Ford -- even though no one in my family, to my knowledge, had ever seen, nor ever intended to see, this movie.

But when Abe saw that we had Blade Runner on LaserDisc, he sensed a great opportunity. See, Blade Runner has always been one of Abe's favorite movies. And since I owned a copy on a premium format, he could get me to fall in love with it while getting to watch it on the big screen with crisp visuals and killer surround sound.

So we made plans to watch it. We fired up the LaserDisc player, and within ten minutes, I was fast asleep. Abe doesn't give up so easily. We tried again. Several times, in fact. And every time the lights dimmed, it was nighty night for me.

It wasn't deliberate. It was simply my teenage body's involuntary response to Blade Runner. Abe thought he was showing me a masterpiece, but my central nervous system clearly felt differently about the film.

That was more or less where matters stood until this summer, as artificial intelligence was moving ever more rapidly from the realm of science fiction into virtually every corner of our lives that I found myself wondering whether I had spent all those years sleeping through a movie whose questions the rest of us were only now waking up to.

So I tried one more time, and this time I stayed awake. And, much as it pains me to admit this publicly, Abe was right: Blade Runner is actually pretty great. Some things are classics for a reason. Beyond stunning visuals, memorable performances, and a killer score, Blade Runner was ahead of its time in many ways. Its questions about created beings now feel uncannily relevant to our age of artificial intelligence. The film imagines a world in which a powerful corporation has created super strong and highly intelligent bioengineered beings called replicants. The corporation that manufactures them even boasts that they are "more human than human."

But in the world of the film, replicants are not regarded as people. They are products -- created to perform dangerous or degrading labor, forbidden from living freely, and destroyed when they become inconvenient -- and the law enforcement specialists tasked with hunting down replicants who have escaped or stepped out of line are called blade runners. Because replicants are not considered human, killing them is not called killing; it is called "retirement."

The trouble is that replicants have become almost impossible to distinguish from biological human beings. They look, sound, and act human in almost every way. So blade runners administer a test to determine whether a suspect is a replicant or a human being. The test measures involuntary emotional responses -- changes in breathing, pulse, and the movement of the eyes -- while posing questions designed to determine whether they react to another creature's suffering.

At first, the film seems to suggest that what distinguishes biological human beings from replicants is that the former possess inherent empathy while the latter do not. But eventually, we see replicants forming attachments, grieving, fearing death, and longing for freedom, while many of the biological human beings remain cold and instrumental, treating sentient creatures as tools, property, and disposable threats.

Our world is not yet the world of Blade Runner, but the movie’s central anxiety no longer feels like science fiction. Machines can now speak with remarkable fluency; they can write, translate, analyze, code, compose music, generate images and video, assist doctors and surgeons, and respond to our fears with language that sounds compassionate. They can perform many tasks we once assumed belonged to human beings alone, and they are improving quickly.

Taken together, these developments raise a question far greater than whether any particular technology is useful or dangerous: What does it mean to be human when machines can increasingly perform tasks we long regarded as distinctively human, sometimes faster or better than we can? 

Even if you have no particular interest in artificial intelligence, I hope you’ll stay with me, because what is at stake concerns us all -- not only what it means to be human in the age of AI, but the ways this technology is already reshaping our shared life. People are understandably worried about its effects on their jobs, privacy, mental health, children’s education, and ability to trust what they see and hear. The data centers that power AI consume vast and rapidly growing amounts of electricity and water; their energy use contributes to climate change, while their demands can strain local resources, often in communities that had little voice in deciding where those facilities would be built. AI is already used for mass surveillance and is accelerating the development of autonomous weapons, while a small number of powerful companies make choices that may shape billions of lives with little public oversight.

And yet the usefulness and potential of AI – its capacity to improve lives and benefit humanity – are equally real. I don’t think using a tool to make life easier is inherently dehumanizing; I use a car even though God gave me legs, and I am not spiritually corrupted every time I use spellcheck. And in any case, unplugging the machines — or at least unplugging ourselves from them — is no longer a practical option for most of us; the technology is already deeply embedded in modern life.

This summer, I read a delightful – if not also a bit disturbing – book by technology journalist Joanna Stern called I Am Not a Robot: My Year Using AI To Do (Almost) Everything. Stern spent a year outsourcing virtually every aspect of her life to AI, and then wrote about the experience for our edification and entertainment. In one particularly memorable chapter, she describes creating an artificial-intelligence avatar trained to imitate her voice and professional style, and then sending it to conduct interviews over Zoom. The AI version of Joanna did not embarrass itself; in fact, it asked genuinely perceptive questions. But that success troubled her more than failure might have.

Interviewing, for Stern, had never been merely a way to extract information from another person. It was a relationship in which connections were formed, surprises emerged, and ideas were born. The machine could produce what looked like a successful interview, but the quality of its output could not measure everything that mattered about the encounter. It could reproduce the product of a conversation, but not the encounter itself.

We might be tempted to conclude that relationship, then, is the answer -- that whatever else machines may be able to do, only a human being can genuinely relate to another.

Maybe. Or maybe the technology just hasn’t gotten there yet. And if we define human uniqueness by whatever machines cannot do, we risk tying human worth to a shrinking set of capacities. Scholar Leif Weatherby calls this “remainder humanism”: each time machines acquire a capacity we once regarded as uniquely human -- calculation, language, creativity -- we retreat to another, defending an ever-shrinking patch of ground against technology’s advance.

But perhaps this entire search for what remains uniquely human begins from the wrong premise. What if human worth does not depend upon identifying some capacity that a machine can’t imitate or surpass?

Jewish tradition starts in a different place. It begins with the Torah’s assertion that every human being is created b’tzelem Elohim -- in the image of God.

Biblical scholars point out that, in the ancient world, kings were sometimes described as bearing the image of a god, and that a king would place his own image throughout his realm to represent his presence and authority. Genesis radically democratizes that royal status: not one ruler, not one class, and not one people, but every human being bears God’s image on earth.

Whatever sacredness inheres in being human belongs equally to every human being, so that no one bears the image of God more than anyone else.

Human dignity is not a prize for winning a competition. It is not awarded for intelligence, productivity, independence, virtue, or empathy, and it does not disappear when our capacities do. Before we accomplish anything, before we prove anything, before we are even aware of ourselves at all, our tradition insists that we are already infinitely, irreducibly, and equally worthy, each and every one of us a reflection of God’s own image.

But what exactly does it mean to be created in the image of a God who has no body or form?

Maimonides, the great medieval philosopher and jurist, attempted to answer this question more than eight hundred years ago. In his philosophical opus The Guide of the Perplexed, he argued that since God has no body, tzelem Elohim cannot refer to physical resemblance. We resemble God not in body, but in mind: our capacity, however limited, to seek and recognize truth reflects God’s all-encompassing knowledge.

But artificial intelligence complicates Maimonides’ answer. If our intellect is the aspect of us that reflects God, what happens when machines appear just as intelligent as we are, or even more capable of sorting fact from falsehood, reasoning through complex questions, and solving complicated problems?

Yet near the conclusion of The Guide of the Perplexed, Maimonides makes clear that intellectual apprehension is not an end in itself. Knowledge of God should lead us to imitate God’s ways by embodying godly qualities such as hesed, mishpat, and tzedakah -- love, justice, and righteousness. To bear God’s image, then, is not merely to know what is true, but to care as God cares about the welfare of others and the world as a whole and to partner with the Divine in advancing a loving and just world.

In Pirkei Avot, the great ancient sage Hillel teaches: “Uv'makom she-ein anashim, hishtadel lihyot ish / In a place where there are no human beings, strive to be human."

In a sense, this teaching seems to contradict the principle that every person is created in God’s image. Tzelem Elohim means that our humanity is something innate, irreducible, and common to us all, while Hillel's teaching implies that our humanity is something we must practice. 

So which is it? Is our humanity inherent, or is it something for which we must strive?

It seems to me that these teachings are not contradictory but complementary. Hillel is distinguishing between being human and living humanely. There may be plenty of people in a particular place and still no one willing to tell the truth, defend the vulnerable, accept responsibility, or care. Do not, Hillel is saying, let the absence of humanity around you become an absence of humanity in you. Human dignity is a given, but humane living is a responsibility. We do not strive to possess human dignity. But we must strive to live in accordance with the dignity entrusted to us.

What does it mean to live in accordance with the dignity entrusted to us? This morning’s Torah portion offers an answer:

God has promised Abraham and Sarah that they will become the progenitors of a great nation, but they are unable to have children. Sarah responds by giving her enslaved Egyptian servant, Hagar, to Abraham so that Hagar might bear the child Sarah cannot. Hagar gives birth to Ishmael, who for more than a decade is the only child in Abraham and Sarah’s household.

Then, against all expectations, Sarah gives birth to a son named Isaac. Some time after Isaac is weaned, Sarah comes to regard Ishmael as a threat to her son’s inheritance and future. She tells Abraham, “Cast out that slave-woman and her son, for the son of that slave shall not share in the inheritance with my son Isaac.”

Notice what Sarah’s statement does to Hagar and Ishmael: Isaac is named, but both Hagar and Ishmael become nameless. They are simply “that slave-woman and her son.” Whatever fears may lie behind Sarah’s demand, her language reduces two particular human beings to their social status: an enslaved woman and her son. Once that is all we see of another person, it becomes possible to regard them as disposable.

Abraham is distressed, but after God assures him that Ishmael, too, will become a great nation, he sends Hagar and Ishmael into the wilderness with only bread and a skin of water.

When the water runs out, Ishmael is close to dying of thirst. Out of options, Hagar leaves him beneath a bush and moves a bowshot away so that she will not have to watch her son die.

And then the Torah says: “Vayishma Elohim et kol hana’ar ba’asher hu sham / God heard the voice of the boy where he was.”

Sarah sees Ishmael only as “the son of that slave-woman.” God, however, hears a child in distress. Sarah reduces him to a category; God recognizes him as a person.

This is tzelem Elohim: a person is always more than a function, a category, a prediction, or a test result. Human dignity is not a verdict rendered at the end of an examination; it is axiomatic and fundamental.

Moreover, God’s care for Ishmael doesn’t end with the recognition that his life matters. God opens Hagar’s eyes to a well so that she can give her son water, and a few verses later, the Torah tells us: Vayehi Elohim et hana’ar -- “God was with the boy.”

To be “with” Ishmael is not merely to regard him as worthy in the abstract. It is to be present with him in his vulnerability and act to sustain his life.

And here’s the thing — our tradition doesn’t tell us how God acts merely so that we will possess accurate information about the Divine. Rather, as beings created in God’s image, we learn about God’s qualities in order to emulate them.

And our tradition has us recall this story on Rosh Hashanah, on this day in which we are called to remember who we are and who we are called to be, in order to remind us that to be human is both to be endowed with inherent dignity and charged with profound responsibility. 

To bear God’s image is to possess an innate, infinite, and irreducible worth. 

And it’s not just a description of who we are. To bear God’s image is also a prescription for who we are called to be. To bear God’s image is a calling, a calling to care as God cares -- to recognize that every person possesses the same fundamental worth. And furthermore it is, like God, to hear the suffering of others whoever and wherever they are, to be with and lift up the vulnerable, and to respond whenever human dignity is denied.

The artificial-intelligence systems we have do not care in that sense. They can assign weights, rank options, predict what we are likely to value, and generate stirring language about love, justice, grief, or conviction. But they do not love anyone, mourn anything, or hold any truth dear. If an AI system seems to express a moral commitment, that commitment came from somewhere outside the machine -- from the people who designed it, trained it, constrained it, prompted it, or supplied the data on which it learned. The words may be generated by the system, but anything those words might appear to care about comes from us human beings.

It is therefore also up to us to decide what values these systems should reflect and what purposes they should serve. Because we can care -- because another person’s reality can matter to us and call forth a response -- we must be careful not to allow tools that do not care to determine what deserves our care. No algorithm can take that responsibility from us, and no tool can absolve us of it.

The deepest danger, then, is not that machines will replace or displace humanity. Rather, it’s that they might make us less humane -- allowing what can be measured and optimized to crowd out what must be valued and cared for, and hiding our choices behind systems that appear impersonal and inevitable.

That temptation -- to confuse what a technology can do with what we ought to ask it to do -- reminds me of an old episode of The Simpsons in which it is revealed that Krusty the Klown is actually Jewish and the son of a great rabbi. In a flashback, the Jews of the Lower East Side of Springfield deferentially bring their pressing questions before Krusty’s father, voiced by the great Jackie Mason. 

Finally, one man asks him, “Rabbi, should I buy a Chrysler?”

The rabbi replies, “Could you rephrase that as an ethical question?”

The man pauses. “Um…is it right to buy a Chrysler?”

And then the rabbi answers, “Yes! For great is the car with power steering and Dyna-Flow suspension!”

Rabbi Krustofsky, of course, has not actually answered the ethical question. He has merely converted it back into a question of features and treated the car’s capabilities as a moral verdict. Artificial intelligence presents precisely the same temptation. The fact that a system can perform a task -- even impressively, efficiently, or cheaply -- does not tell us whether we ought to ask it to do so. Just because something is useful does not mean it should be used.

Nor does potential danger make a technology inherently treif. The question is not whether AI is kosher in the abstract, but whether a particular use will help us live more humanely. Will that use help us honor the Divine image in ourselves and others, or make it easier to disregard the dignity of those it affects?

When artificial intelligence helps a doctor identify an illness, enables a person with a disability to communicate more easily, expands access to knowledge, or relieves someone of dangerous or depleting work, it may enlarge our capacity to care.

But when it allows us to simulate attention rather than offer it, replace a relationship that was itself the point, outsource a judgment whose consequences another person must bear, conceal injustice behind technical language, or treat human beings as disposable because a machine can perform some of their tasks more cheaply, then even a highly successful output may represent a profound human failure. 

Artificial intelligence can help us pursue the purposes we choose, but it cannot choose those purposes for us. It can help us carry out our decisions, but it cannot relieve us of responsibility for their consequences. 

But our responsibility does not end with how we use AI in our own lives. We must also decide together what companies may build, how governments and militaries may use this technology, and what protections must be guaranteed to every person. And because the technology and its consequences cross borders, governing it will require durable laws within nations and binding agreements among them. Those choices do not belong to machines. They belong to us.

And our responsibility also does not end with AI. Long before anyone dreamed of artificial intelligence, Hillel recognized that we have always been capable of being biologically human without fulfilling the promise, potential, and purpose of our humanity. AI gives that ancient moral choice a new and especially consequential form, but Hillel’s challenge remains: Uv'makom she-ein anashim, hishtadel lihyot ish / In a place where there are no human beings, strive to be human."

Let us, today and in the year to come, strive to be human -- not because our humanity is something we have to prove, but because it is something we are called to practice.

Let us strive to be human -- to hear other people where they are, to allow their lives and their suffering to matter to us, and to use our tools to extend our care rather than evade the demands that care places upon us.

Let us strive to be human -- to welcome the assistance these technologies can offer without surrendering the responsibility that remains ours: to decide what matters, what is right, and what we owe one another.

Let us strive to be human -- to remember that a task may become obsolete, but a person never does; to refuse to regard anyone as disposable simply because they are no longer useful to us.

Let us, this day, strive to be human, so that we may bring closer the day when those who wield technological power listen to the people and communities whose lives are shaped by their choices; when burdens are not hidden and benefits are shared; and when technological progress is measured by whether it helps human beings flourish.

For on that day, no one’s worth will be measured by intelligence or productivity. Every person will be heard ba’asher hu sham -- where they are -- and we will use our tools to care for one another as God cares for us.

So may it be God’s will. Amen.