A classroom in Los Angeles. A boy in the back row, fifteen, lets his eyes go flat. A page of Talmud sits open in front of him. The page argues about damage and intent, a quarrel older than any court the boy will fear in his life. To him it reads as static. Rabbi Shlomo Einhorn watches the eyes go flat, and what he feels is close to grief.
This is the scene he returns to in his own telling, and it holds the whole man. A teacher does not fear his own death first. He fears the death of the thing in the next mind. The page lives only if the boy carries it. When the eyes go flat the page dies a small death, and the chain that runs back through the boy’s father and his father’s teacher and a thousand years of teachers stops at a bored fifteen-year-old in the back row.
Ernest Becker (1924-1974) built a whole account of human life out of that kind of fear. In The Denial of Death he argues that man is the animal who knows he will die, and that culture is the answer he builds against the knowing. Becker names two terrors. One is the fear of death, the dread of rotting, of going to nothing. The other is the fear of life, the dread of standing out, of carrying the full weight of a separate self. Against both, a culture hands its members a hero system: a set of roles and rules that lets a man feel he counts in a scheme that outlasts him. Becker calls these immortality projects. You beat death by pouring yourself into something that does not die. The nation. The book. The faith. The work.
The rabbi holds an old and self-effacing version of this. His immortality project asks him to disappear. The chain outlasts the man, and the good rabbi wants it that way. Mesorah, the handing down, runs through him the way water runs through a pipe, and his honor lies in carrying it clean and passing it on. Maimonides (1138-1204) did not write to be Maimonides. He wrote so the law would stand after him. A man who gives his life to that does not fear oblivion for himself. He fears it for the thing in his care.
So set the boy’s flat eyes inside that fear and you see why Einhorn fought the way he fought. He built songs for the hard principles. He drew the Talmud’s quarrels next to the plots of Marvel movies. He turned the law into a game where boys argued like medieval rabbis. His colleagues thought he had lost his mind. The boys started to carry the page. That was the win, and it was a win against death.
Then comes the turn that puts his name on a row of web addresses ending in .ai. He sits with a chatbot and asks it to open a hard passage of Maimonides. The answer comes back correct and dead. He reads it the way he once read the boy in the back row. The machine returns the words and carries none of the life. It makes the student’s old mistake. It can say the thing and cannot hold it.
He had standing. Twelve years a dean, twenty-three years at a pulpit. Here a smaller man stops, and a stranger one keeps going. Einhorn does not conclude that the machine has no place near the sacred. He concludes the opposite. In his telling the realization lands all at once. The skills he spent thirty years teaching, the patient questioning, the building of context layer on layer, the hunt for the pattern under the text, these are the skills that open the machine too. The Talmud, he decides, is the training manual for intelligence that the engineers never read. He starts to build. A virtual study hall. A line of ventures with the kind of names a venture man picks. Chavruta, the old paired study where two men sharpen each other over a page, becomes a thing you do with software, and the rabbi becomes a consultant who happens to wear a rabbi’s title.
Watch the words travel as he carries them across that line, because the same word means a different thing in every hero system it lands in, and the whole question lives here.
Take understanding, the word Einhorn leans on hardest. He wants the machine to understand, not regurgitate, and he means something precise by it, though he never says what.
A heart surgeon means one thing. Understanding is the map worn so deep in the hands that they move ahead of thought and find the bleed before it shows. The resident knows the textbook. The surgeon knows the body on the table at two in the morning. His hero system runs on the patients who walk out and the residents who carry his technique after him, and death stands in the room every time he cuts. For him understanding keeps a man alive tonight.
A Carthusian in his cell means the reverse. For him understanding means surrender. He does not master the verse. The verse masters him. He chews one line for an hour and lets it read him. The slowness is the gate. His hero system holds the world up through prayer no one sees, and the self is the thing he gives away to reach the eternal. Tell him you have made the sacred fast and easy and you have told him you have lost it.
A jazz pianist in the bebop line means a third thing. He learns ten thousand tunes so he can forget them. Understanding sits in the fingers as feel, the knowing of the changes so deep he can leave the melody and trust it to be there. His chain is who taught whom, the lick passed hand to hand in back rooms, the record that outlives the room. Memory for him is the runway. He builds it to leave it.
A trial lawyer means a fourth. Understanding is anticipation, the witness read before he speaks, the jury’s faces, the question that lands because the file lives in the lawyer’s bones. His hero system is the verdict and the record, the win that stands after the client is gone. Memory and care serve combat. The text is a weapon.
A founder in the same Los Angeles world Einhorn now courts means a fifth thing, and his meaning sits closest to the surface of the .ai pages. Understanding is edge, the thesis that moves before the market sees it, the insight a man can defend in a room full of money. He hears the rabbi’s pitch and thinks distribution, moat, the size of the market. Same word. Opposite weight.
One more man belongs here, and the rabbi would know him as kin across every wall of language and faith. A griot in West Africa, or a reciter of the old Icelandic sagas, a man who carries the dead in his mouth. For him to forget a name is to kill an ancestor a second time. His understanding is custody. He holds the line of the living who hold the line of the dead, and he fears one death only, the death of the chain. Sit him beside the rabbi and the two need no translator. They fear the same thing. They guard the same fire.
That kinship is the key to reading Einhorn with the care he has earned. When he asks whether a machine can understand Torah he is not making the founder’s claim or the surgeon’s. He is making the griot’s. He is asking whether a machine can join the line of the living who carry the dead. That is not a novelty pitch. That is the oldest fear a rabbi owns, dressed in a consultant’s clothes. He is not selling a gadget. He is trying to keep the page alive in one more vessel before the eyes in the back row go flat for good.
Now the hard part, because truth asks for it. The thing Einhorn wants from the machine is the thing he saw it cannot give. He read its answer about Maimonides and named it dead. Understanding, not regurgitation, was the line he drew, and the machine fell on the wrong side of it by his own account. His hope rests on the vessel becoming what he watched it fail to be. A surgeon does not hand the scalpel to a tool he just called blind. The rabbi hands the chain to a thing he called soulless and trusts that his method will wake the soul in it. Maybe it will carry the words. Whether it can carry the life is the question.
A second cost follows. Becker would name it the fear of life. The rabbi’s old hero system asked him to vanish into the chain. The new one puts his face and his name on the door, RabbiEinhorn.ai, the man as the brand. To step out of the institution and stand alone under your own name takes nerve, and it carries the danger Becker warned of, that the immortality project starts to serve the self it was built to dissolve. The chain runs through the pipe best when no one can see the pipe. A pipe with a logo is a different thing.
Rabbi Einhorn worships the unbroken line, the handing down that outlasts every hand it passes through. He fears the small death in the back row, the page that dies in a bored mind, the chain that stops at his watch. And the rescue now wears his own face in a way some of his peers tell him to avoid.
None of that makes him a fraud. It makes him a teacher who would rather try a strange thing than stand at the front of the room and watch the eyes go flat. A man who loves the fire that much will reach for any lamp. The reaching is the honor. The lamp is the question.
The People He Defends
David Pinsof, David Sears, and Martie Haselton call their account Alliance Theory, and it makes a hard claim. Political belief systems do not grow from abstract values like equality or authority or tradition. They grow from the structure of who stands with whom. A man’s values track his allies and his rivals. When he argues for a principle, he argues for the people the principle protects.
Run Rabbi Shlomo Einhorn through that lens and a generous portrait comes out, not a cynical one, because Alliance Theory describes a function, and a function can be honorable. Einhorn does the work of raising a coalition’s morale and standing. His allies are the observant, the day-school families, the men and women who keep Shabbat and kashrut and sense that the prestige culture around them files those commitments under quaint. His rivals are less other people than a secular elite that codes Orthodox life as backward, narrow, a thing a bright young man grows out of. Einhorn stands between his people and that judgment. A man who does that for his own is doing loving work.
Alliance Theory holds that we support our allies with a set of tilts the authors call propagandistic biases. We magnify our allies’ grievances and shrink their faults. We assign their wins to virtue and their losses to circumstance. We work to make third parties see our side as the moral one. None of this needs a liar. The theory’s sharpest point is that sincerity is the proof of loyalty. A partisan who shaded the truth on purpose would be a mercenary. The true ally believes.
Einhorn’s gift is the raising of commitment. He makes belonging feel like strength. The young Jew who half hides his observance in a secular room, who reads his own tradition as a list of restrictions he carries out of habit, hears Einhorn and reads it instead as a calling. That shift in feeling holds a coalition together. Alliance Theory predicts quiet defection when members feel embarrassed and thinly attached. Einhorn closes that exit by turning attachment into honor rather than burden.
His second move inverts the rival’s frame. The secular world calls the tradition confused. Einhorn turns the charge around. The confusion sits outside. The tradition knows what it does. This is the boundary work Alliance Theory describes, the sharpening of the line between allies and rivals by feeling more than by rule. Each side in a conflict paints its own as reasonable and the other as lost. Einhorn paints with confidence, and his people stand straighter for it.
His third move raises the stakes. He frames the observant life as a high and serious thing, a calling worth a whole life, not a style one picks among others. High stakes bind the committed and push off the lukewarm. That is the trade his manner makes. It deepens the loyal and thins the marginal.
Take a word like pride. Inside the coalition Einhorn serves, pride in the tradition reads as backbone, the refusal to be ashamed of what your grandparents kept under worse pressure than yours. Carry the same word into the rival’s room and it reads as chauvinism, a closed mind, a people who think too well of themselves. The word holds steady on the page and bends in the world. Alliance Theory says this is the rule, not the exception. The people under a value decide what it defends.
What Einhorn does not do tells you as much as what he does. He does not spend his hours translating Orthodoxy into secular moral language for the doubters at the edge. He does not soften the norms to keep a marginal member from walking. He does not build the slow, dull routines that hold an institution steady across decades. He mobilizes. He does not consolidate. The work he does best and the work he leaves to others are two trades, not one.
A coalition under pressure needs more than one kind of man. Rabbi Yitzchok Adlerstein fortifies the same tradition from the intellect, supplying arguments and tending the long quarrel with the rival culture at its border. Rabbi Yosef Kanefsky works the other margin, lowering the cost of staying for those who might otherwise leave, easing edges so the doubtful keep a foot inside. Einhorn raises the stakes where they lower them. He makes the tradition feel like a calling, and a calling inspires some and loses others. Three men, three functions, one alliance. Alliance Theory expects the division of labor.
His recent turn toward artificial intelligence reads through the same lens. Einhorn now calls himself a consultant, registers addresses that end in .ai, and tells a wide audience that the Talmud is the training manual the engineers never read. Strip the novelty and the alliance logic shows. Silicon Valley holds the prestige currency of the moment. To set the sacred text of his people beside that currency, to claim the tradition trained the very thing the rival elite prizes most, raises the coalition’s standing in the eyes of the third parties who decide what counts as serious. The cure for low status is borrowed status. Einhorn borrows it and carries it home.
The manner carries a cost, and Alliance Theory points at it without scorn. A man who runs on intensity raises commitment fast and strains the routines built for the long haul. The same heat that binds the devoted might wear out a system made for predictability and slow formation. Such men often move rather than settle. Einhorn spent twelve years as a dean and twenty-three at a pulpit, and then he moved. The theory does not tell us why any single tenure ends, and I will not guess. It tells us that mobilizers spend down what consolidators build, that a coalition needs both, and that the two rarely live in the same man.
Hold the portrait at arm’s length and three things stand clear. Einhorn serves the observant Jew who feels his life shamed by the room he walks into, and he serves him by turning that life into honor. He serves him through the ordinary tilts of any loyal ally, the magnified grievance, the inverted charge, the borrowed prestige, none of which need a dishonest man and all of which Alliance Theory counts as the marks of belonging. And he pays for the gift with the thing the gift cannot do, the patient keeping of an institution, which falls to quieter men.
Alliance Theory would file his sincerity as a loyalty signal. The label sounds cold until you sit with it. A man who believes every word he says to his people, who lifts their heads in a culture that wants them lowered, is the kind of ally a pressed people would thank God for. The theory names the function. It does not lower the man.
This morning, The Washington Post published an investigation that should unsettle anyone who cares about the architecture of online influence. Over fifteen months, reporters analyzed 1,435 hours of live streams from white nationalist Nick Fuentes, using AI to scan 2.6 million video frames. What they found was not just grotesque. It was highly organized. Roughly 11,000 donors sent Fuentes nearly $900,000 in superchat payments — digital gifts that flash a donor’s name on screen while the streamer reads it aloud during a live broadcast.
Sit with that for a moment.
After the payment processor takes its cut, he still clears a massive sum. The top 500 donors account for nearly half the total. Around the streams sits an entire auxiliary ecosystem: merchandise, paid inner circles, clip-makers, and a network of loyalists slicing long broadcasts into viral fragments for wider distribution. One TikTok account pushed a single clip past four million views.
That is not random virality. That is infrastructure.
And that distinction matters.
Because the lesson here is not that hate is innovative. The lesson is that hate has been willing to operationalize human needs that more decent corners of the internet have treated vaguely: belonging, recognition, ritual, status, mission, shared language, visible participation, repeatable giving.
The ideology is evil.
The machine is effective.
And unless we are willing to separate those two things analytically, we will keep losing valuable ground to people who understand something simple: communities do not scale on content alone. They scale on rails.
Read the details of the story and your first instinct is revulsion. Mine too. But if you stop at revulsion, you miss the deeper point. The internet has already run the experiment. It has already shown that thousands of people will repeatedly fund a worldview when that worldview gives them identity, participation, and liturgy.
The core move is sound, and not new. Separate the ideology from the infrastructure. Grant that belonging, recognition, ritual, and visible giving are human, not the property of cruelty. He states it well, and the line about decent people confusing niceness with structure is true and worth saying. A lot of good work loses because it trusts the message to travel on its own. He also speaks to two rooms at once, the Talmud line and the Yoda line, and that is a real skill, not a trick.
Truth first, though. The essay enacts the thing it studies. It opens on a hate-streamer and a $900,000 figure, rides the moral urgency of a Washington Post exposé, runs eight numbered tools to hold your attention, and lands on a product. Those are his own rails. The shocking enemy pulls you in, and the destination is Mallacore and a paying client, Jerry Joyner. He discloses the client, which is rare and to his credit. The disclosure makes the piece honest. It does not make it analysis. By the third “that matters” and the third round of credit for Jerry, you are reading advertising.
The deeper trouble sits in his own quotations. He cites Yoda. The dark side is quicker, easier, more seductive. That line argues against him. If the rails carry corrosion better than they carry gratitude, the form is not neutral. His first tool makes the point for me. Build identity around refusal, find an enemy, even if the enemy is “a pattern.” He concedes the engine needs an enemy, then tries to launder it. Contempt produces a compulsive return that gratitude does not. He wants the same dopamine with the opposite spiritual effect and assumes the swap comes free. The whole piece rests on that unexamined assumption.
Then the evidence. He builds a scalable playbook out of a single extreme case. We see the Fuentes operation that worked. We do not see the thousands who built the same rails and drew nothing. The concentration he admires, top 500 donors supplying half the money, is a story about a small fanatical core, not a model that ports to earned optimism. One proof of concept is not a blueprint. I take the Post figures as he reports them; my read does not turn on the numbers.
Now the empathy. The man opens with midah tovah merubah and he means it. He believes good has greater overflow, and he is trying to make goodness compete against a machine built for cruelty, which is harder and more honorable than waiting for goodness to win on its own. His charge that the well-meaning have been structurally unserious is humane and right in the main. And he hands you eight usable things. That is a teacher’s generosity, not a guru’s fog.
Last spring he wrote as a rabbi who had expanded his toolkit, Torah at the center, AI as the new vessel for an old love. Now he writes as the founder and CEO of an AI firm, one Talmud line up top and business mechanics filling the rest. The center of gravity moved from the text to the company. Many men make that move. I name it, and attach no verdict.
So: a real idea, half right, delivered in a sermon-shaped sales document that performs its own thesis, weakened most by an author who quotes the case against himself. The kindest true thing I can say is that the impulse is good and the vehicle is compromised, and he was honest enough to tell you so himself.
‘The Updated “Signs AI Wrote This” List’ (Mar. 16, 2026)
What is damning is the cluster. Over-symmetry combined with over-polish combined with fake insight combined with generic abstraction combined with predictable rhythm. When those five things appear together, readers feel it before they can explain it.
The Tells, In Order of How Much They Hurt You
Writing in threes. “Strong, sound, and scalable.” “Clear, grounded, and practical.” “Simple, powerful, effective.” The triad creates instant rhetorical symmetry, which is exactly why it reads as manufactured. Humans stumble into threes occasionally. AI reaches for them as a default rhythm engine.Fake profundity transitions. “This is the part I want to tell you.” “Here’s what most people miss.” “Let that sink in.” “The truth is.” These phrases create the sensation of revelation without the substance behind it. The tell is not the phrase itself. The tell is that the sentence following it rarely justifies the buildup.
Over-clean contrast structures. One well-placed contrast is a legitimate rhetorical move. Repeated contrasts feel templated. “Less about X, more about Y.” “From X to Y.” “The goal is not X, it is Y.” When a piece uses this move four or five times, the reader stops feeling persuaded and starts feeling processed.
TED-talk in text form. Short sentence fragments stacked for drama. “The kind that shows up. In hard moments. When no one is watching.” This was a genuinely effective technique when it was rare. Now it is one of the most recognizable AI-coded aesthetics in public writing…
Stake. Say something specific enough to be wrong. Say something that has a cost attached to it. Write as if someone you respect will read it and disagree, and you are saying it anyway.
The underlying shift in 2026 is that audiences are no longer spotting vocabulary. They are spotting template logic. The question is no longer “did a machine write this?” The question is “did a person actually think this through?” Those two questions have started to feel the same.
Most of the list is right. Writing in threes, the fake-profundity transition, the over-clean contrast, the paragraph that always lands on a tidy thesis, low burstiness, no idiolect, the keynote ending that means nothing. He has a good ear. He reads enough to have heard these patterns and named them, and the closing turn is the sharpest thing here: the tell moved from vocabulary to template logic, and the real question is whether a man thought something through.
Now the problem. The piece is a specimen of the thing it hunts. It writes in threes throughout. It runs the contrast structure it warns about in almost every entry. The tell is not the phrase, the tell is the sentence after it. The question is no longer this, the question is that. It stacks fragments for drama. Elegant. Empty. Every item ends on a neat takeaway, which is the relentless closure he flags a few lines later. It mocks “X is not about X, it is about Y,” then leans on that frame to close the whole essay.
And the three fixes he prescribes are the three things the piece lacks most. Specificity: no exact street, no precise number, no detail that could come from one man and no other. Stake: nothing here is costly or embarrassing or his own. Idiolect: no pet phrase, no odd syntax, no person behind the wheel. By his own test the essay fails. It could have been generated. Some of it may have been.
That is the real trouble with the genre, and he steps right up to it and stops. Once “asymmetry reads as human” becomes common knowledge, asymmetry turns into a style to fake, and the overlong sentence and the abandoned thought become a new template. He names the edge of this with “template logic,” then backs away. The honest version of his own essay runs about two hundred words: the detector migrated from words to structure, the structure is now gameable too, and what remains is the old question of whether anyone thought. Everything above that point is the filler the point indicts.
The essay he meant to write lives in his last two paragraphs. The rest is the genre he is describing.
A man who makes his living deploying AI wrote the guide to spotting AI in a voice the guide would flag. He gets close to something real about thinking against generating. He does not turn it on his own prose, which is the one place it would have cost him something, and cost is the thing he says good writing carries.
‘The Talmud as AI Model’ (Mar. 12, 2026)
For fifteen years I sat in a beit midrash, watching students wrestle with texts that were written to resist easy answers. The Talmudic sugya is a remarkable intellectual machine: it opens with a bold claim, invites a devastating objection, entertains a minority view that reframes the whole question, and only then issues a practical ruling. Nothing is accepted. Everything is pressure-tested.
When I moved into AI deployment, I kept running into the same problem with language models. Ask a question, get a confident answer. Ask for analysis, get a well-organized opinion. The model would rarely challenge its own first move. It would commit to a frame, elaborate on that frame, and dress up the result with appropriate caveats at the end. Sophisticated-looking. Structurally shallow.
I started wondering whether the sugya could fix that.
This one is the best of the four essays I’ve read so far because the rabbi is writing from the thing he knows. Fifteen years in the beit midrash give him the sugya from the inside, and the mapping he draws is clean and not forced. Claim, objection, reframe from a different premise, ruling with conditions. That shape is real, and laying a prompt over it is a fair and useful move. Where the AI-tells piece had no specificity and no stake, this one has both. The drop-in prompt is concrete. A man could paste it tomorrow and get a better answer on a strategy call than a single pass gives him. Credit where due.
Now the trouble. His account of why it works is shakier than the tool itself. He says hallucination is confidence without friction, and that structural resistance cures it. But the objection the model generates comes from the same engine as the claim. The model can invent a self-generated kushya the same way it invents the first answer. His own step two says cite or simulate evidence. Simulate evidence is an instruction to make things up with better posture. You can get the choreography of rigor with none of the grounding, and that reads more trustworthy than a plain answer while being no truer. The form of contradiction is not the substance of it.
The disanalogy he skips is the one that carries the weight. The sugya’s objections were not improvised. They came from a closed canon of recorded positions, argued inside a tradition that transmitted them. The rigor was downstream of the corpus and the community. The model has the moves and none of the substrate. It performs the beit midrash with no books on the shelf and no men across the table. So the structure can sharpen reasoning, and I think it does, while doing nothing to anchor a claim to fact.
One line deserves a flag. He says the rabbis were not chiefly after truth, only after reasoners who could not be fooled. That is a strong claim about a tradition that prizes emet, and he asserts it because his AI point needs it. The sugya bends to fit the tool. A reader who did not know better would take it as settled. It is not.
And a touch of oversell at the close. Two thousand years, the prompt already written, we just had to see it. Structured adversarial prompting is not new. It runs under other names in the prompt literature. His contribution is the framing, the idiom that makes the technique land for a particular reader. That is worth something. It is not a discovery.
The tool is good and the explanation is off. The prompt helps because it forces the model to surface considerations a single pass skips, not because it stops the model from inventing. Keep the prompt. Drop the claim that it reduces hallucination, or keep it tentative. And notice the pattern across his work. When the Torah carries the load instead of seasoning the top, he gets specific, and the writing gets good.
‘How AI can help you with Mesilas Yesharim’ (Dec. 16, 2025)
The pursuit of ethical perfection in Rabbi Moshe Chaim Luzzatto’s Mesillas Yesharim presents a unique challenge: the Ramchal explicitly states that his work contains no new information for the scholar. The difficulty is not yediah (knowledge) but hargashah (internalization) and chazarah (constant review).
Large Language Models offer an unprecedented opportunity here. While AI cannot possess a soul or perform commandments, its capacity for infinite patience, rigorous logical structuring, and persona simulation makes it an ideal “external cognitive scaffold” for Cheshbon HaNefesh (accounting of the soul).
The prompts that follow are designed to transform passive reading into active, dialogic character refinement. Each prompt can be copied directly into ChatGPT, Claude, or any capable AI assistant.
How to Use These Prompts
• Copy the entire prompt text (everything in the gray boxes) and paste it into your AI chat.• Engage authentically — these work best when you answer honestly, not performatively.
• Use consistently — the Ramchal emphasizes daily practice. Consider scheduling specific prompts for specific times.
• Adapt as needed — these are starting points. Modify the language to fit your life circumstances.
This sits with the sugya piece as his strongest work, and for the same reason. He starts from the text and reads it right. The Ramchal (1707-1746) says outright that he teaches the scholar nothing new, that the gap is not yediah but hargashah and chazarah, and Einhorn builds for that gap and not for information. That is the correct read of Mesillas Yesharim, and most people who quote the sefer miss it. The prompts are concrete and a few are good. The Rationalization Adversary, the machine voicing your strongest yetzer-hara arguments and then taking them apart, maps onto nogea b’davar and ahavat atzmo with real care. The Invisible Man and Credit Swap tests turn lishma into something a man can check. He read the book.
Truth first, though, and the trouble runs deeper here than in the other pieces, because the substrate is supposed to be your own soul.
Mussar runs on friction. The hargashah he names comes from the self wrestling itself, from the reckoning done by the one being reckoned with. Hand the cheshbon hanefesh to a machine with infinite patience that does the structuring for you, and you risk stripping out the friction that makes the work work. You get the report without the reckoning. The flashlight he describes is held by someone else. It lights the room. It does not train your eyes. He waves at the line, no soul, no mitzvot, and then the prompts cross it. The Vigilance Audit has the model deliver the Ramchal’s verdict. The Rationalization prompt has it rule clean or tainted and quote the principle at stake. An LLM invents those quotes the way it invents anything, and now it paskens your moral state. The Ramchal worked inside a mesorah and sent real questions to a rav. The prompt seats a confabulator as posek and mashpia.
Then the flattening, which runs through the whole thing. The mussar arrives dressed as behavioral science and consulting. Behavioral Engineer. Habit stacking. Two-Minute version. Heart MRI. Strategic Risk Analyst. The desire-deconstruction prompt tells the model to strip a craving to its chemistry and clock the crash at thirty minutes, which is Marcus Aurelius and dopamine talk, not the Ramchal. This is the move he makes every time. He translates the sacred into the prestige idiom of the moment. Last spring it was Silicon Valley. Here it is the habit-science shelf at the airport bookstore. The translation buys reach and thins the thing.
The impulse is good. He wants people to walk the path, not just read about it, and the sefer begs for exactly that. Mussar has always borrowed the tools of its day. Rabbi Yisrael Salanter would recognize the instinct if not the device. For a man with no mashpia and no chavruta at 2am, an interrogator of infinite patience might be the thing that gets him to start at all.
But the same trait that makes it accessible makes it thin. The patience, the absence of judgment, the always-on. A mashpia helps because he remembers you across years, because the relationship carries stakes, because he can rebuke you and love you in one breath. The machine forgets, risks nothing, and can do neither. It can scaffold the shape of cheshbon hanefesh. It cannot supply the witness.
‘How AI is Revolutionizing Israel Advocacy’ (Jul. 29, 2025)
Imagine knowing exactly when and where negative sentiment about Israel begins to spike online before it becomes a trending topic. Advanced AI sentiment analysis tools like Sprinklr, Brandwatch, or even custom solutions built on platforms like Twitter’s API can continuously monitor millions of social media posts, news articles, and forum discussions.
These tools go beyond simple keyword tracking. They analyze context, tone, and emotional indicators to provide nuanced insights. For instance, they can distinguish between legitimate policy criticism and antisemitic rhetoric, or identify when discussion about Israel shifts from neutral news coverage to emotionally charged debate.
The practical application is powerful: advocacy teams can receive real-time alerts when sentiment shifts, allowing them to deploy resources strategically. Teams can intervene early with factual information and context, rather than discovering a viral misinformation campaign after it has already spread.
This is the weakest of the six, and the cause sits closest to his heart.
The piece is generic in a way the sugya and Mesillas Yesharim pieces never were. Swap “Israel advocacy” for nonprofit fundraising or real estate marketing and almost nothing changes. He is not writing from a place he knows here. He is listing tools. Sprinklr, Brandwatch, Video Authenticator, IBM Watson, Jasper, DeepL, NodeXL. And the list has aged badly. Microsoft’s Video Authenticator was a minor 2020 release, never a frontline detector. Watson was a fading brand by 2025. He name-checks Twitter’s API as if it were open, when access had been priced out for two years. The roster reads like it came from old training data, not from a man who has run these systems.
This is from July 2025, and it is the purest specimen of the voice he would teach people to detect eight months later. Game-changing. Force multiplier. Lightning speed. The most immediately practical application. The frictionless keynote ending about who will be best positioned. By the checklist he published in March, this piece fails on every line.
Now strip the banner and read what he proposes to build. Sentiment surveillance across millions of posts. Chatbots dropped into WhatsApp groups and Twitter threads to inject corrections. Network maps that find the key accounts and the communities where an idea takes root, so teams can hit “strategic intervention points.” Netflix-style targeting that feeds each man the version of the story he is most likely to swallow. That is an influence operation. It is the same apparatus, rail for rail, that he would admire in the Fuentes machine nine months later and try to invert. He recommended building it before he saw what it was. Automated persuasion and bot-seeded argument corrode the information commons no matter whose hand is on the switch. Doing it for a cause you believe in does not change what it does to the room. The same line applies to every side that builds these tools, which is the point.
Two technical claims deserve a flag. He hands a classifier the job of telling antisemitism from legitimate policy criticism, at scale, as if that line were settled and machine-readable. It is neither, and false positives there carry real cost. And he sells autonomous fact-correction, a bot that cross-checks an image and posts the correct information, as reliable. The same overconfidence about AI runs through all six pieces. The machine is steadier in his telling than in life.
He keeps the human-element caveat, frames AI as helper and not replacement, warns against one-size messaging. People who care about a cause often move slower than the people working against them, and that gap stings. He wants to close it.
But this is the piece where sincerity and craft sit furthest apart. He cares most about this subject and brought the least thinking to it. He reached for the tool list and the marketing voice, and he never stopped to ask whether the machine he was urging people to build is the machine he distrusts everywhere else.
‘Truth, Justice, and the Torah Way: Exploring the Jewish Conception of Justice’ (July 10, 2025)
The libertarian vision of justice resonates deeply with Torah values emphasizing personal freedom and individual responsibility. The Torah repeatedly underscores the dignity of the individual, exemplified vividly in the divine creation narrative: “So God created mankind in His own image” (Bereishit 1:27). The Mishnah (Sanhedrin 4:5) elaborates powerfully, stating that the creation of a single person teaches that each individual holds infinite value, and “one who saves a life, saves an entire world.”
The Rambam (Maimonides), in his Mishneh Torah, champions a system designed to protect individuals from violence and coercion, aligning closely with Nozick’s minimal-state concept. Halachah, Jewish law, notably in tractates such as Baba Kamma and Sanhedrin, meticulously outlines prohibitions against theft, violence, fraud, and invasion of personal property, underscoring a robust protection of individual liberties.
Yet, Torah’s understanding of justice extends beyond mere non-interference. The Jewish tradition continually insists upon an active moral duty towards communal welfare, fundamentally diverging from pure libertarian minimalism.
The move at the center of his essay is the safest one in comparative thought. Set up two poles, libertarian and egalitarian, then land in the wise middle that holds both. Almost any tradition read with sympathy can be made to do this. Swap in Catholic social teaching or Confucianism and the essay changes almost nothing. The synthesis is so wide it cannot be wrong, and a claim that cannot be wrong does not tell you much. He reaches balanced wisdom too fast, which is the moral-neatness tell he would publish a warning about eight months later.
The good parts are real. Tzedakah from tzedek, obligation and not charity, is a true and clarifying point for a lay reader, and the West-versus-Torah contrast lands. The built-in redistribution, Shmita, Yovel, Peah, Leket, is apt. The Rambam’s eight levels with self-sufficiency at the top, correct. When he stays on Jewish ground he is fine.
The trouble starts when he imports the axis. Nozick (1938-2002) and his Anarchy, State, and Utopia name the libertarian pole, but the egalitarian pole stays anonymous. The man he wants is John Rawls (1921-2002), and A Theory of Justice goes unnamed in a piece pitched on a philosophy hook. That is a hole. The deeper problem is that the Torah does not run on the liberty-versus-distribution axis at all. Its categories are covenant, mitzvah, the ownership of all wealth by God. Tzedakah is not liberal egalitarianism with Hebrew roots. It is a commanded obligation that flows from God’s claim on what you hold, a different thing wearing the same coat. He dresses the Torah in Nozick and Rawls so it reads in the seminar room. That is the move he makes in every piece, the sacred translated into the prestige idiom of the hour. Last year Silicon Valley. Here the political-philosophy syllabus.
One citation needs a flag, and I checked it. He credits the Maharal (c. 1512-1609) with Netiv HaTzedek. The Maharal’s ethical work is Netivot Olam, and its charity path is Netiv HaTzedakah. A path of justice sits beside it, the path of din. Netiv HaTzedek is not among them. Worse for his case, the Maharal there does something more pointed than the balance of freedom and solidarity Einhorn assigns him: he ties charity to justice, to din, and sets it apart from chesed. Einhorn flattens that into a warm equilibrium. The citation is plausible and a little off, which is the hazard that runs through all his work.
Here is the through-line that struck me. This July 2025 essay is a field guide to the AI tells he would publish in March 2026. He delves in the second paragraph, the first meme-word on his own later list. Nuanced appears three times. Dynamic equilibrium sits in the middle. The prose stacks adverbs, vividly, powerfully, meticulously, passionately, profoundly. It closes on a benediction about societies worthy of divine blessing, the frictionless ending he would name. He wrote the specimen, then wrote the diagnosis.
And the bow pulls against the essay. He spends the piece arguing that Torah transcends the libertarian and egalitarian split, then turns and says the American way might reflect that Torah wisdom after all, which folds the distinction he just built.
As a four-minute devar Torah for a lay audience, it’s fine. Judaism does not slot into the American binary, and tzedakah-as-duty is a correction. He does not bend Torah to a program. He lands on both-and, where the tradition sits. But the essay is safe, the frame is borrowed, and it carries the voice he would later teach his readers to distrust.
‘Why Positive Psychology Turned Virtue Into Self-Help Snake Oil’ (July 3, 2025)
Positive psychology didn’t discover virtue, it strip-mined it. When Christopher Peterson and Martin Seligman published their handbook of character strengths, they performed an impressive feat of cultural archaeology. They dug up wisdom from Aristotle to Confucius, catalogued 24 character strengths under six virtues, and handed us a neat little assessment tool.
But something died in translation.
Aristotle spoke of eudaimonia, literally “good spirit,” a flourishing that came from aligning your soul with cosmic truth. The Stoics saw virtue as living in harmony with the rational order of the universe. Eastern traditions connected character development to transcendence of the ego and unity with ultimate reality.
Peterson and Seligman kept the vocabulary but gutted the vision. They gave us Wisdom and Knowledge, Courage, Humanity, Justice, Temperance, and Transcendence, but these became personality traits to optimize, not pathways to truth.
This is virtue after the lobotomy.
The Zombie Shuffle
Walk into any corporate wellness program and you’ll see them: zombie virtues shuffling through PowerPoint presentations. “If your top strength is Kindness, try three random acts of compassion this week!” “Got Curiosity? Take an online course!” “Scored high on Gratitude? Start a journal!”
These aren’t virtues anymore, they’re life hacks with ancient names.
Real virtue was never about self-improvement. It was about self-transcendence. The ancients understood something we’ve forgotten: you can’t cultivate authentic character by focusing on yourself. Virtue only makes sense when it points beyond the self to something larger, truer, more enduring.
The fact/value split killed that understanding.
