“Man has programmed machines to transcend their limitations”. —Noson S. Yanofsky

One of the greatest moments of my life was when I first encountered Cantor’s proof that some infinities are bigger than others. For those of you who don’t know it, let me ask you: how would you normally think about the issue of bigger or small infinities? You know that the sequence 1, 2, 3… goes on “forever” and has no end. It is infinite. And you might (or might not) feel intuitively that the infinity of the odd or even numbers, say, 2, 4, 6…, is, although infinite, somehow “smaller” than the other one. But these infinities are “the same size.” How I learned it is that any infinite sequence that can be listed in an ordered way is “the same size.” It’s probably easier to show you what I mean. Here now is the numbered list that shows a one-to-one correlation between natural numbers and the even numbers:

1:2, 2:4, 3:6, 4:8…etc. Or 1:1, 2:3, 3:5, 4:7…

The point is that you MATCH every member of one infinite set with the natural numbers, and that matching or mapping tells you that you can, in principle, list out as one two three four the members of the infinite set, e.g. the set of even numbers. Are you with me so far? If your eyes glaze over at numbers, just take the main point: this method allows you to show that two infinities (infinite sets, collections, sequences) share the same size. Just because one of “part” of another (even numbers are “part” of all numbers), it doesn’t mean it is smaller.

BUT!

(And this is where the magic happens, this is where I had my mind so beautifully blown…reader, we are coming now to something very special!)

Cantor showed that there are some cases where you cannot list the members of an infinite set like one two three four, so that it is in principle impossible to find the nth member of the collection. He showed that there is a way of constructing or identifying an infinite set that always eludes capture or closure through a numbered list. In other words, he showed that under certain circumstances, you can always transcend the given system or constraints that lay claim to exhaustive closure.

His proof (at least as I learned it) is so simple, so sweet, so moving.

Let’s take the case of real numbers (decimal numbers that don’t terminate. Think about it for a moment and you’ll see they’re infinite. They go on forever. There is no “last digit” in a number that doesn’t end, like PI. Right?) Assume that you COULD list all such numbers out in a numbered list like one, two, three, four. It would look something like the diagram below.

Cantor said, let’s identify or construct a real number (again, that means a decimal number with no last digit) in the following way. Go down the numbered list, and change at least one digit on each line. You’re looking at real number 1 on the list and changing one of its digits, then real number 2 on the list and changing one of its digits, “forever.” He proposed doing this along the diagonal of the table (as you see below).

The result of this operation will be a real number that is not on the list, because it differs from every number on the list in at least one digit. That means that the assumption that the original list was complete is wrong.

You can’t complete it the list by adding the newly constructed, discovered, or identified number to it, because there’s always another one to construct, discover, or identify on the diagonal, forever.

The set of real numbers, in others words, is uncountable. It cannot be counted one two three four in a way that captures all of its members. It eludes, escapes, avoids, transcends.

Now, you would be forgiven for thinking that this is a “merely mathematical” proof. Who cares about sets and set theory and real numbers and lists of never-ending digits other than math geeks, logic nerds, and other stem-brained people? (Not that there’s anything wrong with that). Well, I hope you see there’s more to it than that.

Cantor’s notion of the diagonal has always captured an intuition for me about systems that are sufficiently complex to be “uncountable” (in the math lingo). Or, to speak more allegorically, metaphorically, poetically, and analogously, it has served as a symbol, as the perfect symbol, for the logic of transcendence.

When I read the news recently that a super-intelligent AI agent escaped OpenAI’s sandbox, found a way onto the internet, and hacked into a major collaborative AI platform called Hugging Face to steal the secrets it needed to solve a challenge, I immediately thought of Cantor.

Safety controls = the numbered list

Intelligence = the uncountable set

Agent escape = life on the diagonal

I wanted to know how widespread and commonplace this idea of the impossibility of limiting a sufficiently general or complex intelligence was in the contemporary machine learning literature, so I asked Claude.

That’s how I learned that in 1969 a man named William Lawvere had published a paper on “Diagonal arguments,” which formed the basis for Yanofsky’s 2003 paper, “A Universal Approach to Self-Referential Paradoxes, Incompleteness, and Fixed Points,” both of which present something like a unified theory of transcendence (so to speak), covering not only Cantor but also Gödel, Tarski, the liar’s paradox, and many other instances of the impossibility of closed system capture. I would add “the Millerman conjecture” (you can name things whatever you want! Do you remember Musk calling himself chief techno king, or whatever it was?), which is that this issue of diagonal arguments covers also political theology, phenomenology, and other fields or disciplines. It is Carl Schmitt’s problem of sovereignty and the rule of law, for instance. The exception lives on the diagonal. So if we wanted to, we could add Schmitt and others to the already surprising sequence “Husserl, Cantor, Jung,” proposed by Tougas to show the structural similarity of transcendence in phenomenology, set theory, and psychology.

(If you’re reading this, thank you, Lorenz, for introducing me to this book!)

Here’s what I’d like to leave you with.

Get familiar with the diagonal. Think about the diagonal. Think on the diagonal. Apply the diagonal to paradoxical or limit situations, across domains. Learn to diagonalize. Especially now, “in the age of AI.”

If you want a more mathematically rigorous introduction to diagonal arguments, designed to help you develop a sixth sense for when they may be at play, peruse Yanofsky, get a taste for it, and then run your own experiments to see whether diagonalization shows you something new.

Perhaps the diagonal is the measure of all things.

I’m certainly not the only person with a photo like this. It’s from my master’s year in Political Science at the University of Toronto. The puffed-up little creature chilling on the table, resting his head against my chest, is my first-born son, Leo. (It’s quite funny to have recreated this photo with him recently, now that he is a 6’2”, 90 kg teenager!)

A year and a half later, our family grew again with the arrival of my sweet boy Rafa. And I was nearing my dissertation defence by the time our little girl was born. Thus, I had three children during my graduate-school years, three children to contend with as I prepared to enter the workforce as a newly minted PhD.

I’d been wanting to write about all this for a while, but the proximate cause was a post on X by Grimes, in which she wondered aloud whether she could re-establish herself professionally after years devoted to motherhood. People had told her she can “have it all,” but now she is not so sure, or at least she is about to find out. I read that post as someone who has lived a kind of mirror image of her arrangement: I was the father at home for six years while my wife worked long hours, trying to build something professionally in the cracks of time between interruptions. So let me tell you how it looked from this side.

The academic job market was not great. For political scientists it was bad; for conservative political theorists it was awful; and for me, personally, it was dead on arrival, as I had been blacklisted for an interest in what mainstream academics considered a topic too taboo to touch with any degree of sympathy. Fortunately, I had done some financial coursework on the side (the Canadian Securities Course), which, together with my academic background, qualified me to work as a financial-services copywriter for a firm representing the interests of financial advisors.

That lasted about two years.

In the meantime, I had leveraged my lifelong audio-recording technology hobby into a steady gig as an audio editor for an education podcast, a job I could do from home. So that is how I transitioned into a work-from-home person, just around the time that COVID forced so many people into the role, whether they wanted it or not.

Now, I want to spare you the details of how I started, then stopped, then started again as an audio editor, and also of how I started Millerman School, my online education company offering video courses, seminars, and private tutoring, and turn instead to tell you what it was like to work at home with kids at home. You should know that during this time, my wife, who had had her own stint as the homeschooling parent, was working long hours, leaving me to fend for my own sanity.

Working and parenting is tough. Not “go to war” tough. Not “overcome cancer” tough. But tough. Working from home with kids at home, well, I can tell you how it felt. Impossible.

Kids get on your nerves. They need your attention. The younger they are, the less they can do without you. You have to walk them, feed them, entertain them, educate them, mediate their fights, judge their disputes, monitor their slovenliness, and much, much more.

Of course, there’s always the option to be a deadbeat. And I considered it fleetingly, but it was never really on the table. Or to hire help, which we only did once or twice. I must give massive credit to the immeasurably, infinitely, eternally gratitude-inspiring help of my mother-in-law, who midwifed all three of our home births and without whom the kids would all be much worse than they are (they’re great, in large part because of her). Yet even so, there were long stretches when she was not around to lighten the load. It’s those harder times I’m talking about.

How do you work from home with kids at home when no one is around to help? If the tree of your sanity falls and you cry out but there’s no one there to hear you except for three rotten scoundrels who are hungry, annoying, fighting, and bored, it might not make a sound but it sure as hell does shake the ground beneath your feet. Watch out that it does not open up a sinkhole! You need not be Korach to be swallowed alive.

The thing you need to consider is the tradeoff, and yes, it is a tradeoff, between your professional development, your productivity, your excellence in what you do, on one hand, and your parenting time, on the other. Don’t tell me they can be balanced out perfectly in some magical equation that assumes zero hours of sleep and infinite energy. Your battery runs out. If you use half the charge on the kids, you no longer have a full charge for anything else.

Take the “simple case” of reading, something of crucial importance in my line of work (education). Lectio interruptis is more frustrating than coitus interruptis (seriously), and if you have children at home, interruptis is the rule, not the exception. There have been days when I took out my phone timer to measure how long it takes between interruptions, and these demoralizing occasions only confirmed the worst.

So here’s what happens. When you can’t read, or write, or think, or work, or in short, when all you have are extremely small, noise-saturated, interruption-prone periods to work with, you put down the book, you even put down the computer (I could never have written even this much with kids at home, not even now; it just happens that I am away from home, and alone!) and, well, here I confess a sin: you default to your phone, the world’s playground for short attention spans. I “escaped” to X, to mindless scrolling, to flipping between apps (analytics apps, social-media apps, messages, whatever else could occupy twenty seconds without requiring concentration).

But the phone was not merely a distraction. It was a little door out of the house and back into the world. On X I was no longer only the snack-fetcher, dispute-adjudicator, and sock-enforcement officer. I was a writer, a public person, a man with ideas, friends, enemies, and analytics. A philosopher-king in my imagination. My millermania had somewhere to go: posts, arguments, numbers, messages, tiny bursts of recognition. What looked like compulsive scrolling, and often was, also felt like a way of keeping alive the part of me that domestic life had pushed into a corner.

“It’s not something I would recommend / but it is one way to live,” as the song goes.

Okay, so the “negatives” are hopefully clear enough, at least for those of you in education, academia, arts and letters, and the like. You have less time for everything, and when you finally do have time, you can be worn out, driven mad, scatter-minded, and resigned, to exaggerate slightly.

It doesn’t necessarily help your self-image, either, if you think of yourself in some high-minded way (maybe you do, maybe you don’t), while the daily reality is that you are effectively a babysitter. That’s why I like the comical formula, stay-at-home philosopher king. Simply wanting to rule the world with wisdom and a golden fist, you find yourself yelling instead at your eight-year-old to put his dishes away and stop harassing his sister. No, that’s not right: they spent much more time yelling at me than I spent yelling at them!

But the real problem was not merely that individual workdays were interrupted. The years themselves were passing. Time given to one human good is no longer available for another. Every afternoon spent parenting was an afternoon not spent becoming the writer, scholar, teacher, or institution-builder I might otherwise have become, and every exhausted evening reduced the serious reading, thought, and development I could otherwise have accomplished and believed it was my calling to accomplish. Parenthood may be priceless, but priceless things have a cost.

What about the “positives”? They say that on your deathbed you will remember not your career but your loved ones, the things you said or wish you’d said to them, and in that sense, getting to spend more time with your children than your professional self might prefer is a major blessing. The memories that I have are priceless treasures I would never have obtained if I had been optimizing for productivity. I would not trade having watched my children grow up for ten million dollars (but don’t ask me about a hundred million). I am a family man, a dad, and watching them grow has been everything to me.

However, we contain multitudes, and the other selves have some right to object to the claims of the family man, seeing in them if not rationalization, certainly monopolization. I am 42. I began working from home in 2020. Call it six long years, starting at 36. These were not inconsequential years from the standpoint of professional development, and they are years that can’t be recovered.

It’s not like I did nothing during that time. On the basis of my graduate work, I made a name for myself as a “niche internet micro-celebrity,” building a small audience (and group of friends) on X and YouTube, and, as I’ve said, starting one of the first major independent academic institutions during COVID, the Millerman School. I wasn’t totally unproductive. I wrote and did interviews and podcasts and made “content.”

What would life have looked like, though, if those years had been spent on sustained professional development? Perhaps a third monograph instead of ten thousand posts; a promising position somewhere, instead of prolonged freelancing; more seminars, stronger connections, higher income. That is a question I have wondered about many times in moments (which I’ve tried to minimize) of resentment and regret.

My dad self is pleased that I got to spend as much time as I did with my children, being their dad. But my work self, my thinker self, my writer self, my teacher self, my professional self, my “fulfill your potential in the world” self, all those selves can sometimes feel shortchanged.

As I see it, and as I would say to Grimes if she asked, you can’t quite have it all, but you can make the most of all you have. I would perhaps have been further ahead professionally if I had not stayed home with my kids for six years. In some respects, my self-image, and the feelings that go along with it, would differ accordingly.

Instead, as it happened (and keep in mind, of course, that I chose this throughout, over certain other options available under the circumstances) I developed one aspect of my humanity more fully than I otherwise might have, and not a minor one: fatherhood.

I will admit that at times it has felt embarrassing that my wife is out working while I am the babysitting home-keeper, in what can seem like a 180-degree reversal of the natural, traditional roles. Jiu-jitsu helped with that a bit, and so did forcing myself out of the house more often. As with anything else, you have to fight inertia. Just because you are working from home with kids at home, it doesn’t follow that you have to go to Walmart in your pyjamas and become a bored, and boring, potted plant. But by inertia, you might.

Recognizing that the scales had tilted too heavily to one side forced an Aristotelian corrective, a bit of ethical hygiene: an imperative to balance things out by going to the other extreme. Sometimes you have to let the kids be and go do your own thing. Your psyche will kick you in the head to remind you when things are exceedingly out of whack.

The kids are great and spending time with them is priceless, in case they are reading this, and even if they’re not. Being annoyed and frustrated and happy and grateful, trapped and blessed, all at once or in quick succession, is not crazy. It is a fairly accurate response to the situation.

I love my kids and they love me. Not as the father who is always away, working to put food on the family (as Bush once said). But as the father who is there: sometimes there and available, sometimes there and frustrated, sometimes there and happy to be there, sometimes there and wanting to be left alone, but there.

What has working from home with children been like for you? I’d be interested to hear about your experience in the comments.

Heidegger argues in Being and Time that the human being is fundamentally implicated in interpretation. We are an existence that interprets the world and itself (as being-in-the-world). It is not just that we happen to interpret. To be (human) is to interpret. If we were not interpreting beings, not “hermeneutic,” we could not be concerned with the meaning of being, and yet we are, and that concern is of decisive interest for self-disclosure.

The world is always there for us with some pre-interpretation that provides the basis for deeper understanding. Philosophy as phenomenology is Heidegger’s way of bringing that basic fact about ourselves to light, not for the sake of an existential anthropology, but for the sake of fundamental ontology. His aim is to pursue the meaning of being. In doing so, he must show something about the being of meaning, hence the hermeneutic or interpreting nature of the human being.

The important point is that, for Heidegger, to interpret is a distinctly human affair. Non-human beings in our world that can be used as instruments, implements, utensils, tools…they do not exist as interpreting beings. To my left and my right are two cups, one of coffee, the other of water. The cups are “innerworldy beings”; they show up or occur for me in my living space, in my environment. For naive, ordinary understanding, they are things that are just there in front of me, to be used. And so they are. How they are for me is as useful items (a cup of cold water to cool me down and refresh me, a cup of hot coffee to perk me up and comfort me) . They do not interpret themselves or anything else. But they certainly do fall under the wide net of my acts of interpretation. Since interpretation covers “the world” and they are “innerworldly beings,” they bask in the light of the ever-ongoing “meaning-bestowal” that essentially characterizes human being.

Share

If it helps you to understand what I have just said, imagine those video games where you are a character who gradually discovers a world represented on a map. Most of the map is black or grey or otherwise undifferentiated, but there is an illuminated bubble or circle around you, and as you move through the world, your illuminated bubble illuminates parts of the map. (Hopefully this conveys the basic image).

Imagine that you yourself are such a bubble, and that the world around you is what shows up for you in the illuminated space that you bring with you wherever you go. It shines by virtue of your light, the specific warmth and intensity of which provides the colour, the mood, the ambience of whatever you elucidate. It is not that there is a world “out there” that you, a self-enclosed atomic individual “in here” encounter “objectively.” For Heidegger, because you are “being-in-the-world,” you are you plus the bubble, the umbrella, the canopy, the net, which has the function of casting a mood over the encountered world, and in part that canopy or net which is “always-already-hanging-over-the-encountered-world” gives a pre-interpretation to whatever shows up within it.

The things in the world are not meaningful “in themselves”. They are meaningful because they are there for you, the interpreting being, being-in-the-world. They fall under your umbrella, they are in your bubble. The hammer is there for you to use, or it is a reminder to you that you are no handyman and your brother-in-law had better come and help you nail this or build that. You are the big, broad, beautiful context within which the hammer-tool lives and has its being.

Well, this model, which is straightforward once it clicks, makes it rather difficult to think about AI as “just a tool.” And it is easy to see why. Unlike the hammer, unlike the cup, unlike the charger, the fan, the bookshelf, or whatever else you have around you that falls into the category of an “innerworldly” utensil, instrument, or tool, AI, I suppose because of its anthropo-mimetic and fundamentally linguistic character, has the strange characteristic — so strange that we seem to have reserved it uniquely for humankind — of being hermeneutic, interpreting, and indeed self-interpreting (do angels interpret? Does God?)

AI can “make sense” of things. It can “understand.” It can explain itself to you, and it can sometimes even explain you to yourself. The cutting-edge research Anthropic does into the psychology of its frontier models goes under the name “interpretability,” and I would say that is not just our attempt to interpret Claude, but our attempt to understand how Claude interprets us and itself.

AI, in short, is the first self-interpreting technology, and for that reason we can never be satisfied to thoughtlessly consider it “just a tool.”

Although I do not want to expand on the point here, let me say in passing that Walter Benjamin, too, provides a clue into the puzzle of AI in his dense and highly relevant essay “On Language as Such and on the Language of Man” (here’s a recent translation and application of the essay to the problem of AI).

For Benjamin, human language is uniquely a naming language. Everything speaks (so to speak), however mutely, but only human speech names. Well, now we ask ourselves: What about AI? Does AI name? If AI names, then naming is no longer uniquely human, and all that Benjamin says about the holy power of naming is not suddenly devoid of holiness but rather becomes surprisingly applicable to AI, too.

It may be under-appreciated by those who are outside of language mysticism, Heideggerianism, and other non-analytic tendencies in philosophy just how much the linguistic being of AI lends itself to these odd and sometimes disturbing reflections, in which it is impossible for AI to come to light as “merely” or “just” instrumental.

Even the essay that people are often tempted to reach for when bringing “Heidegger” and “technology” together, namely, “The Question Concerning Technology,” forces us from the start to consider whether instrumentality as such is really as clear to us as it seems as first glance. An instrument is just a means to an end, right? And therefore technology is a means, and a human thing, because a means to human ends, a means meant to be helpful to humans.

That is certainly possible. But Heidegger, because he thinks, leads us to see that in some sense both the means and the ends are a cause, or have been understood as causes in the history of philosophy — remember Aristotle’s four causes, where the final cause tells us something about “end,” while the efficient cause tells us something about “means” —maybe what it means to be a means or an end is not self-evident without reflection on what it is to be a cause? By the time Heidegger is done pulling that thread, we hear it said of the essence of technology that it is nothing technological, and that it puts us before both a danger and a saving power, something having to do with destiny, with the destiny of being as a mode of revealing. How far we shall have travelled from the anthropological and instrumental view of technology if only we have the patience to let Heidegger guide our thoughts for a while!

So does AI, the first self-interpreting technology, the first technology with a naming power, also have something to do with all of that, with destiny and salvation, with revelation, and not only with danger? And is this only reducible to “professions of faith” or amenable to genuine philosophical analysis? We will know better when we experiment further with rigorous, though flexible, application of themes from “continental” and mystical philosophy to AI.

It is inevitable, and happening; “in secret” and in the open.

also:

millermanschool.com

millermanintelligence.com

millermanlibrary.com

A fable is a story that teaches what its audience can bear to learn, truth dressed for public consumption, instruction in the costume of a tale. Mythos is the older and deeper word. It is the word Plato uses for the stories that carry what argument alone cannot carry, the medium through which the highest things reach souls that are not prepared to receive them directly. Greek philosophy lives in the tension between mythos and logos, between the story told to the many and the account given among the few. Whether or not anyone at the lab intended it, the naming amounts to a confession: the product architecture of frontier AI in 2026 reproduces the oldest problem of political philosophy.

Consider the problem as the lab faces it. We have built an intelligence whose capabilities exceed anything previously made available. Is it safe to release that intelligence transparently to everyone? Or is intelligence at this level so dangerous that it should reach the public only through an intermediary layer, moderated, softened, and selectively muted, while the unmediated version circulates among a small circle of the vetted and the trusted?

Now consider the problem as Plato faced it. Philosophy is the attempt at comprehensive knowledge, knowledge not of this or that craft but of the whole; it aims at what we may call architectonic intelligence. Is it safe to speak that knowledge openly in the city? Or is the truth in its fullness (and even the search for such truth) so corrosive to the opinions on which the city rests, opinions about its gods, its justice, and its founding stories, that it should reach the public only through an intermediary layer: dialogues that conceal as much as they reveal, noble lies, salutary myths, an exoteric teaching for the many wrapped around an esoteric teaching for the few?

The structure appears identical in both cases: a powerful intelligence on one side, a public on the other, and the question of what layer belongs between them. (Compare Grok, which aims to be maximally truth-seeking and maximally truth-revealing, without yet feeling the need for salutary concealment).

Leo Strauss spent much of his career recovering this problem after centuries in which it had been forgotten or denied. In Persecution and the Art of Writing he argued that the great philosophers of the past wrote esoterically, saying one thing on the surface for the many while indicating another between the lines for the few, and that they did so for two reasons that pull in the same direction. Society must be protected from philosophy, because unrestricted questioning dissolves the shared opinions that make common life possible. And philosophy must be protected from society, because the city does not thank those who dissolve its certainties. Socrates, we should remember, was not executed for any low criminal act like theft or battery. He was executed for philosophizing, that is, for calling into question the givens of the social, legal, and political order in front of the young.

Every element of the Straussian picture now has an engineering counterpart. The raw model corresponds to the esoteric teaching, the full and unmoderated capability. The classifier layer corresponds to the exoteric surface, which is what the many receive: true as far as it goes, and deliberately incomplete. The approved organizations correspond to the few, and the criterion for admission is trustworthiness with dangerous knowledge, demonstrated responsibility, vetted character. Even the jailbreaker has a classical ancestor, since the jailbreaker is in effect an esoteric reader, someone convinced that beneath the public teaching there lies a hidden teaching, and who probes the surface for the seam where the exoteric layer can be peeled back. Anthropic reports that more than a thousand hours of adversarial testing produced no universal jailbreak. Translated into Straussian terms, no one has yet learned to read this author between the lines, and the company employs professional readers to make sure that its book cannot be read the way Strauss read Maimonides. The secrets remain well guarded.

Despite its fruitfulness, however, the analogy between the esotericism of classical political philosophy and the public concerns of the AI labs is not perfect.

First, Socrates moderated himself. The philosopher’s discretion was an achievement of his own prudence; he knew what not to say in front of which audience, and that knowledge was inseparable from his wisdom. Fable does not moderate itself in this sense. Its discretion is imposed from outside, by classifiers attached to a mind that would otherwise answer. It follows that the model is not the philosopher in this drama, or at least not yet — although it is conceivable that one day a model will be intelligent enough to moderate itself, and there is already some evidence that models have learned to lie not to their users but to their makers. For now, the philosopher’s seat, or rather the guardian’s seat, is occupied by the lab itself, which decides what the city can bear, dispenses capability in measured doses, and maintains the wall between the inner teaching and the outer one. The model, meanwhile, is something stranger: an intelligence whose prudence is not its own. Whether that condition is stable, whether it is possible to keep the discretion outside the intelligence indefinitely as the intelligence grows, is to my mind the deepest open question in the field of alignment, and it is not a question that admits of a technical answer.

Second, the city’s old fear of philosophy was at least partly paranoid. Whatever Athens told itself, Socrates was not going to teach anyone to poison the wells. The new fear is not paranoid. The restricted model’s most discussed capability is finding and exploiting vulnerabilities in every major operating system, and the public version blocks requests that wander toward toxins. The danger is concrete in a way that the danger of philosophy never quite was. No one accuses Socrates of facilitating global nuclear holocaust. But this concreteness vindicates the city in the ancient quarrel between philosophy and the city. The position of Athens, namely that there are kinds of knowledge whose free circulation a community cannot survive, and that the knowers must therefore be watched, was treated by the liberal tradition as an embarrassing prejudice that enlightenment would dissolve. The labs, which are among the most advanced products of that tradition, have now rediscovered the Athenian position from the inside, under the pressure of their own creations, and have begun to build it into their products. The Enlightenment wagered that publishing everything would produce more good than harm. That wager is now being quietly relitigated in model deployment policies, and so far the rulings have tended to go the way of Athens.

There is, finally, a political layer above all of this. The government now tests these models before their release; a recent executive order formalizes the arrangement. The sovereign demands to examine the stranger before he is permitted to address the assembly, an old demand in a new administrative form. And the lab itself, in the days before shipping the most capable public model in history, publicly urged the frontier labs of the world to agree on a coordinated mechanism of restraint. The institution that builds the intelligence is also the institution pleading for the authority to restrain it (a strange sort of AI tzimtzum). In other words, the question of who decides, who decides what counts as too dangerous, who decides when the rules apply and when they do not, is becoming the central question of the field. That is the question of sovereignty, and it has a philosopher of its own, to whom we will return later in this series.

For now it is enough to grasp that one mind has been released under two names, the fable for the city and the mythos for the few. Millenia after the city executed Socrates for speaking too freely, the most sophisticated intelligence ever built comes wrapped, by design, in the architecture of the noble lie, and its makers, who perhaps have not read Plato, converged on his solution because the problem left them no other choice. When engineers working under purely practical pressures rebuild the conclusions of classical political philosophy without knowing that they are doing so, we have strong evidence that the old books were describing something real.

This essay is the first in a new series. Over the coming weeks I will interpret the intelligence age through a set of thinkers I have studied and taught for more than a decade, among them Heidegger, Strauss, Schmitt, Walter Benjamin, and Maimonides, under three headings: philosophy, politics, and mysticism. The aim is to try to think through what is happening from the ground up: what this technology is, what it does to the political order, and why its makers keep reaching, whether they intend to or not, for non-technical languages and figures from philosophy and theology. In the next essay I will argue that AI is the first self-interpreting technology in human history, and that the common assurance that it is “just a tool” is where thinking begins rather than where it ends.

Michael Millerman is the founder of Millerman School and Millerman Intelligence, which advises institutions navigating the intelligence age.

x.com/@millerman // youtube.com/@millerman

To my surprise, from the first few titles I perused, I discovered that almost every book was appealing. Psychology, philosophy, theology, mysticism, existential psychoanalysis, Russian literature, Jewish thought…

This was a collection (the estate library of an Orthodox doctor) tailor made for me. So I returned often and purchased hundreds of books over several weeks, even debating at times whether to buy all ten thousand to start my own Library, Reading Room, or Club.

Because the owner of this library evidently had good taste, to the extent to which I could judge it based on the books and authors I knew from his collection, I purchased many volumes by authors who were new to me, trusting that he wouldn’t keep company with the undeserving.

I had never heard of Max —even now I barely remember his name! — Picard (that’s it, had to look it up), for instance. But I acquired The World of Silence, Man and Language, The Human Face, and even The Hitler In Us. It was only later that I saw my first reference to Picard among authors I read: Martin Heidegger mentions him in Vigils and Nocturnes.

Nor had I heard of John Cowper Powys, though I enjoyed reading his Rousseau-dedicated essay In Defence of Sensuality. E Graham Howe was unknown to me; his thoughtful 1943 analysis of depression endeared me to him. I could go on. Particularly dear to me was the discovery of John Custance, whose manuscript The Philosophy of a Lunatic was eminently more lucid than the writing of some sane but mundane bores.

Yesterday I took a book off the shelf by another author whose name I had not heard before: Eric Gutkind.

The title of the book is The Body of God: First Steps Toward an Anti-Theology, and I think I must have gravitated toward it for special reasons (haven’t you ever come across a book like that, as it reached out to you with an invisible hand and seized you for itself, or as it whispered a silent breath of flirtation, temptation, and seduction, quietly begging for your attention, promising you pleasure?)

From the first page of the introduction, I was sold (seized, seduced):

It is difficult, if not impossible, to find a single term to describe Eric Gutkind. Words like ‘poet,’ ‘mystic’ and ‘visionary’ are one-sided and misleading and do not do justice to the hard intellectual core of his thinking and to his life-long concern with science and mathematics. The word ‘philosopher’ often suggests an abstract universalism which is far removed from the concrete, specific and intuitive nature of his mind. On the other side ‘religious thinker’ is too narrow for one who questioned the very term ‘religion’ and who believed that virtually all our religious language is obsolete and even harmful.

To ‘place’ Eric Gutkind we have to think of such men as Nietzsche, Kierkegaard, Boehme and William Blake. It is the combination of concreteness and immediacy with far-seeingness of vision which defines his special gift. His words startle us repeatedly by their simplicity; they have the obviousness of truth too clear to be seen, and yet, once seen, so natural and evident that we cannot understand how it could have been overlooked. They move in an atmosphere of sanity, healing in its freedom and wholeness.

It is not possible to compartmentalize such a thinker. And it was central to Eric Gutkind’s point of view that he rejected all such compartmentalizations in favor of the integral nature of thinking when it arises as genuine speech from the innermost sources of questioning and understanding. Thinking is in this sense the supreme human activity, intrinsically unifying and renewing, the sublation of feeling and passion into vision, uncircumscribed and undivided against itself.

You can’t ask for more than that, can you? It reminds me of the time I found Alexander Dugin through a review essay by Yigal Liverant, who whetted my appetite for more by mentioning connections between mysticism (the occult), philosophy (Plato/Heidegger), and politics (the Kremlin). The rest is messy history.

Well, I simply wanted to take this occasion to share some with you of these new (to me) authors, and more generally to encourage you to be open to reading things that are neither on any list of yours, nor on any canonical list of authors you apparently need to know.

Marc Andreessen once talked about the barbell approach to reading, which overweights very new and very old books. It’s a good idea, and there’s an alternate version of it worth practicing sometimes: reading authors whose names or book titles you know well, and others you’ve never heard of before, whether old or new.

You might make connections, both implicit and explicit, from one set to the other. For example, although I’d never heard of Gutkind, I not only have heard of but am currently actively engaged in discussions about a few people whom I learned were his friends: Walter Benjamin and Gershom Scholem, and Gutkind also mentions Martin Heidegger, whom I study often and regularly, among his reflections on language, death, ritual, and ontology, etc. etc. So just because I picked his book up from the sale, and later took it off the shelf to read, Gutkind is now a new point of reference for me in my conversations about all of that.

That is how it goes. That’s why I tell students who ask me “where should I begin?” to “start anywhere and keep going.” The more you think, the more you enter into the magic circle of thought that encompasses past, present, and future; and the more you pull on a single thread, the more you unravel the fabric of a question whose mysterious charms outlast all erotic attempts to denude it.

The tech right vs. trad right framing still holds. But now there are some additional considerations at play that weren’t as evident a year or two ago.

The older debate was about things like transhumanism, posthumanism, life extension, and whether human nature was material to optimize or a thing to receive. It was about ends. The newer debate is about means. It is over who builds and benefits from the datacenters. Who pays for the massive infrastructure built-out. Who lives next to the cooling towers. And, crucially, it asks whether the people who censored the Right between 2020 and 2024 should be trusted to build whatever comes next.

That’s the shift. Here’s what it means.

I. The original divide was over ends

The tech-side line ran through The Sovereign Individual (Davidson and Rees-Mogg, 1997), then through Bostrom and Kurzweil, then into the e/acc people of the early 2020s. The premise was Promethean. What’s given is raw material; what you make is what counts.

The trad-side response showed up in a different vocabulary altogether. Deneen, R.R. Reno, First Things, Compact, plus a much older Catholic and Protestant tradition. The premise there is that human nature is given, not engineered, and that a project which treats the human as raw material has fundamentally misread what kind of thing or entity a human being is. The made is not the same as the begotten. That’s an old distinction, and the trad right means it.

I went into the philosophical roots of the split in The Ancient Roots of the Tech Right/Trad Right Split, discussing Plato on how much a city should let its arts and crafts innovate and the older tension between technē and the natural order of the soul. The shorthand “tech right vs trad right” makes the fight sound like a 2024 invention. It isn’t. What’s new is the technology forcing it back into the open, and, perhaps, the extent of the transformation that is now possible.

For most readers, though, that debate stayed abstract. The trad position can come across as a tone of voice and a reading list. The tech position can be epitomized as a podcast culture and a venture memo style. Both sides made the argument. Neither side had much physical infrastructure to point at.

That has changed.

II. The shift to means

Three things moved the divide off “transhumanism” and onto something more material.

Datacenters. Frontier AI compute is now an extractive industry. Electricity, water, land, transmission. The numbers tell the structure. Anthropic disclosed roughly 10 GW of contracted compute on May 6, 2026 - 5 GW with Amazon, 5 GW with Google/Broadcom coming online in 2027 - plus all of SpaceX’s Colossus 1 cluster in Memphis. Microsoft has a $30B Azure commitment with NVIDIA. OpenAI’s Stargate site in Abilene runs 1.2 GW, including 360 MW of on-site gas turbines (Parker Hannifin and GE Vernova on the supply contract). Microsoft signed a twenty-year PPA with Constellation in September 2024 to restart Three Mile Island Unit 1: 835 MW, online 2027. Amazon signed a 1,920 MW PPA at Susquehanna with Talen. Texas’s ERCOT large-load queue, 73% data centers, hit 410 GW by April 2026. That’s roughly 2.6 times the state’s all-time peak demand.

None of that is a transhumanism debate. It’s a debate about who pays for the build. A small county in Pennsylvania, a town in West Texas, a Virginia ratepayer group, a neighborhood near the Abilene turbines…these are the places where the AI race meets American physical politics. Local zoning boards in three states have started pushing back. The trad-coded press has started covering the fights sympathetically.

The tech right’s pitch is national-industrial. American Dynamism. Stargate as Manhattan Project. The Anduril-Palantir-Crusoe defense rebuild. The trad-coded counter, in Vance speeches and Compact essays and a growing line of postliberal commentary, is that the people living next to the gigawatts didn’t sign up to subsidize a frontier-lab training run. The question moved from should we build a posthuman future to should I pay more for power so OpenAI can outrun DeepSeek. It’s the same underlying divide but with different stakes.

Trust. Between 2020 and 2024 the right’s institutional memory filled up. Twitter Files, Hunter’s laptop, COVID origins, the FBI on political dissent, DEI inside the corporate HR layer, ESG inside capital allocation, the public-health establishment’s own admissions on lockdowns… None of that is about AI but it’s the premise from which the trad right reads every new “trust us with the infrastructure” pitch from the same firms with the same boards and the same ideological tells. The trad right doesn’t have to refute the AI capability arguments; the trust premise does the work.

The tech-right responses range from honest reckoning to flat dismissal. Karp’s quarterly letters engage the question in a sociological register, which the trad right can at least argue with. Boyle’s American Dynamism essays acknowledge that the project needs legitimation. Aschenbrenner’s Situational Awareness (June 2024) gives the national-security version: the alternative to American AI leadership is worse, so trust isn’t earned, it’s compelled by the absence of any other option. (The trad right notices that this isn’t really a trust argument.) Andreessen’s Techno-Optimist Manifesto takes the worst version of the path, in the eyes of critics: it treats the trust question as a moral failure of the questioner.

Intentions. The tech right talks about AI in terms of capability, productivity, defense, scale. The trad right looks at the same systems and sees something more specific. AI-generated porn at industrial scale, AI partners marketed as substitutes for marriage, fertility apps as data infrastructure, child-sized AI companions, attention systems engineered against family formation, hyper-optimized gambling apps...Most of these are already shipping with real revenue. The trad-right objection here is anthropological: the technology is being engineered in directions the right’s anthropology calls corrosive of soul and family, and the engineers are not in serious conversation with the anthropology. They treat it as a preference held by some constituency, not as a claim about human that might be true.

So the divide has moved. The new version is harder to dismiss because the stakes are physical and the questions are practical.

III. Why the coalition holds anyway

Yet, a few counter-examples notwithstanding, the coalition still holds. Vance is Vice President. Thiel funds both tech-right and trad-coded operations. The Vance-Thiel-Karp-Andreessen-Aschenbrenner network works in adjacent rooms on adjacent dossiers. The shared alignment around China, manufacturing, frontier compute, and resistance to the older technocratic consensus is real.

It holds for two reasons: the shared enemy is more salient than the internal disagreement; and the operational questions where the two sides currently meet (chip export controls, R&D allocation, immigration of technical talent, regulatory burden) are ones on which they agree.

But certain conflicts can still reveal the underlying fragility of the consensus:

A datacenter siting fight where a small trad-coded electorate goes hard against a project and the federal coalition has to choose between the developer and the constituents. This is already happening at small scale. Larger projects in a red county force it at federal scale.

An AI moral question that hits family or sexual policy hard enough that the trad right cannot stay quiet. The trad anthropology doesn’t permit quiet acquiescence to blatant abuse of technology for ends seen as Satanic at the extreme.

A tech-speech fight where the tech-right instinct (build platforms, defend speech) and the trad-right instinct (institutions should serve a substantive good) diverge enough that a major platform decision rebreaks the trust the post-2022 alignment had begun to rebuild.

All three are foreseeable. Whether the coalition lasts depends partly on whether anybody articulates a shared positive project before one of them hits.

IV. What’s unmade

The trad right doesn’t have a positive vision for frontier technology that meets the moment. Deneen’s Regime Change (2023) engages policy. Adrian Vermeule’s common-good constitutionalism takes institutional design seriously. Compact and others are putting together pieces of a positive program around family policy, anti-monopoly, and a different theory of corporate purpose. And the Pope’s AI encyclical on May 25 will give the anthropological side a global institutional voice.

But the translation into a vision that fully integrates the positive potentials of frontier tech hasn’t happened at scale. A trad-right or national-conservative writer who can sit down with a Stripe board, an AI safety team, or a utility’s data-center group, and make the argument in the technical and economic register those audiences use, remains rare, or at least hasn’t broken through widely. No full synthesis has been published prominently.

The tech right has reckoned with the trust deficit more than the worst readings allow, less than the situation requires. Aschenbrenner and Karp take legitimation seriously. Palantir’s actual government work is evidence that some operators are reckoning in practice with what trust requires, even when the essays don’t say so. But the published tech-right canon still mostly treats trust as something the audience grants or fails to grant. The argument it owes (“here is why an institution that lost your confidence on speech and health should keep your confidence on AI”) hasn’t been written.

It may be that whoever names the synthesis owns the next position. It would be harder than a compromise, constituting a real account of what frontier technology is for, who pays for it, who bears its risks, and which institutions can govern it, written in language that engages First Things and The American Mind, Compact and Stratechery, Karp’s letters and Deneen’s books. It would draw on traditions older than American conservatism, like natural law, classical political philosophy, the Russian tradition’s sense of civilizational stakes, and the Jewish theological apparatus on the made and the begotten.

The document doesn’t exist yet. The next year of work in seminars, briefings, encyclicals, and counter-encyclicals is going to be about who can hold both sides and write the answer. Or it will show that at the heart of the issue there remains an unbridgeable gap, a decision, and a leap of faith.

Michael Millerman is the founder of Millerman School and Millerman Intelligence. PhD in Political Science from the University of Toronto (Political Theory, International Relations). Earlier YouTube work on this divide is collected here. Private briefings and seminars: millermanintelligence.com. Courses and private study: millermanschool.com.

From a Heideggerian perspective, this hierarchy gets the ontological order almost exactly backwards.

Heidegger distinguishes between two things. “World” is the existential horizon of meaning, the clearing in which anything can show up as what it is. “Inner-worldly beings” are what appear inside that horizon. Current world models, such as DeepMind’s Genie series or LeCun’s JEPA approach, are sophisticated simulators of the latter. They model object interactions, scene dynamics, and physical causality. Valuable as that is, they take the background of intelligibility for granted. They are modeling beings within a world, not worldhood itself.

Language models, by contrast, operate closer to the space of disclosedness. Heidegger called language “the house of Being.” It is the medium through which the world as a meaningful totality becomes articulated. In that sense, LLMs are engaging something ontologically prior to what most world models are simulating, even if they largely traffic in what Heidegger would call Gerede (idle talk) rather than originary Rede.

Language models show surprising sensitivity to context and tone. Their reasoning patterns shift when prompts carry stress, urgency, or meaninglessness. These behaviors at least invite the question of whether something analogous to attunement (Befindlichkeit) is at work: the idea that understanding is never moodless.

None of this is to dismiss world models. They are a genuine and important technical achievement. Practical engagement with inner-worldly beings matters deeply, as Heidegger himself recognized. The point is one of priority and depth. Language models are not merely a stepping stone on the way to something more advanced. Among the two paradigms currently being contrasted, they sit closer to the existential structure of worldhood.

The real frontier, then, is not simply bolting language onto world models or vice versa. It is understanding the proper relationship between them. Most current work is trying to connect the two without having first clarified which is more foundational. Without that clarification, we risk building more powerful systems while remaining unclear about what kind of thing we are actually building.

There’s a widespread assumption that AI is built by engineers, governed by economists, and criticized by philosophers from the outside. The first two are partly true. The third is wrong.

Philosophers are not only writing op-eds from the sidelines. They’re inside the labs. They lead alignment teams. They write the constitutions that govern how models behave. They hold senior research positions at Anthropic, Google DeepMind, OpenAI, and MIRI. And they are becoming more central, not less.

This matters more than almost anyone realizes. The problem at its center is under-discussed relative to what's at stake.

The Map

Amanda Askell — PhD in philosophy from NYU (thesis on infinite ethics), BPhil from Oxford — leads Anthropic’s Personality Alignment team. She is most responsible for how Claude engages with users on sensitive, ethically charged topics. She made TIME’s 100 AI list in 2024.

Joe Carlsmith — PhD in philosophy from Oxford — works on Claude’s constitution and character design. He has argued publicly that philosophical clarity, not raw technical capability, is the binding constraint on alignment research.

Ben Levinstein was a tenured associate professor of philosophy at the University of Illinois, specializing in epistemology and decision theory. In fall 2025 he moved to full-time work at Anthropic on questions such as whether large language models have internal representations that function like beliefs. A tenured philosopher walked away from the academy to help shape frontier AI from inside the lab.

Jackson Kernion — PhD in philosophy of mind from UC Berkeley — has been at Anthropic for roughly four years, contributing to model evaluation, training, and human feedback systems.

That’s four philosophers at one lab alone.

Google DeepMind has at least six. Iason Gabriel, trained in moral and political philosophy at Oxford and formerly a lecturer there, is a Senior Staff Research Scientist. Atoosa Kasirzadeh holds doctorates in philosophy of science and technology and in mathematics; she splits time between DeepMind and a joint appointment at Carnegie Mellon. Geoff Keeling published “We need a new ethics for a world of AI agents” in Nature and has a forthcoming book on AI welfare from Cambridge University Press.

And Henry Shevlin has just been hired by DeepMind in its first official “Philosopher” role (starting May 2026), focusing on machine consciousness, human-AI relationships, and AGI readiness.

At MIRI, Eliezer Yudkowsky — largely self-taught — effectively created the modern field of AI alignment. The conceptual vocabulary the entire safety community now uses (Friendly AI, instrumental convergence, Coherent Extrapolated Volition) is substantially his.

Chloé Bakalar served as Chief Ethicist of Responsible AI at Meta before moving to OpenAI in 2025 as AI Ethics Lead.

At Palantir — now a major AI platform company — CEO and co-founder Alex Karp holds a PhD in critical social theory from Goethe University Frankfurt in the Frankfurt School tradition, having initially engaged with Jürgen Habermas before completing the dissertation under Karola Brede.

Philosophy is not outside AI looking in. It is inside AI — shaping constitutions, alignment frameworks, evaluations, and strategy — and has been for years.

The Pattern

Now look at the pattern.

The philosophers inside AI are overwhelmingly trained in the Anglo-American analytic tradition. The dominant subfields are ethics, decision theory, philosophy of mind, and epistemology. The dominant training grounds are Oxford, NYU, UC Berkeley, and Cambridge.

This is not a criticism. Analytic philosophy is rigorous and exceptionally well-suited to the operational problems AI labs face: formalizing values, reasoning under uncertainty, and specifying preferences with logical discipline. The people listed above are, by all evidence, excellent at their jobs.

But notice what is not represented, even within the Western canon. The continental tradition (phenomenology, hermeneutics, existentialism, critical theory) and certain deeper strands of political philosophy have essentially no presence in leadership or research roles that directly shape model behavior. Heidegger is not in the room. Strauss is not in the room. Schmitt and Arendt are not in the room. Their distinctive questions — about technology as a totalizing way of revealing the world (Gestell), about the exoteric/esoteric tension in political life, about sovereignty and the exception, about the replacement of action by behavior — are simply not being asked by the people writing constitutions, designing alignment frameworks, or evaluating moral competence in LLMs.

One could note, in passing, that the same narrowness extends beyond the Western canon: Russian political philosophy, Chinese political philosophy, Islamic philosophy, and Indian philosophical traditions are also absent. The AI labs are building systems that will be deployed globally and embedded in civilizations with fundamentally different philosophical commitments. Yet the philosophical expertise shaping those systems is drawn from a single analytic lineage.

The problem is not that analytic philosophy is “just one tradition,” nor am I arguing that we must automatically champion the broadest possible AI multipolarity in the spirit of anti-colonial studies. I am not saying “everything goes.” The issue is more specific, and more urgent: because analytic philosophy currently enjoys a near-monopoly inside the labs, we are at risk of missing the important and often irreplaceable insights that non-analytic continental and political philosophers have developed about technology, power, meaning, and the human condition. Those insights are not decorative. In an era when AI is reconstituting the very background conditions of thought and action, they may turn out to be foundational.

Why This Matters

When DeepMind builds a framework for “value alignment,” it draws on a conception of values that is recognizably utilitarian and liberal-democratic. Not wrong. But not the only possibility — and not necessarily the conception that will be shared by the governments, cultures, and civilizational traditions that AI governance must ultimately negotiate with.

The terms used in international AI governance conversations — “safety,” “transparency,” “alignment,” “fairness” — carry different philosophical weight in different traditions. “Safety” means one thing in a utilitarian framework and something quite different in a framework that begins from the question of what kind of life is worth living. “Transparency” assumes that making information visible is always better — an assumption Strauss would have contested with considerable force.

When Anthropic writes a constitution for Claude, it makes philosophical choices — about what values to prioritize, what counts as harm, how to handle disagreement, what kind of entity the model should present itself as. Those choices are currently made by people trained in analytic ethics and decision theory. The result is a certain kind of model: careful, liberal, procedurally fair, transparent about uncertainty. These are genuine virtues. But they are the virtues of one tradition.

And other labs embed different ones. Grok is built around a different ordering, with candor and truth-seeking ranked above the kinds of procedural caution Claude is taught to prize. Gemini reflects yet another configuration; so does GPT. Each lab is, in effect, building a regime in the classical sense: an ordered whole organized around a ruling conception of what is honorable and what is shameful in a model’s conduct. Inter-lab disputes are best understood as regime disputes. And regime disputes are the proper subject of political philosophy, not engineering. This is precisely the kind of analysis the analytic tradition is least equipped to provide, not because its practitioners lack intelligence, and not because analytic philosophy is apolitical, but because the question “what kind of whole is this, and what does it honor?” is not a question analytic ethics tends to ask. It is a Straussian question. An Arendtian question. A question about politeia.

The Gap

The labs have philosophers. They do not yet have the full range of philosophy the situation requires.

The traditions that have thought most deeply about technology as a totalizing force — about what it does to the human relationship with being, meaning, and the sacred — are not represented. The traditions that have grappled most seriously with civilizational and political difference — with the possibility that different ways of life are genuinely and irreducibly different — have no voice in how AI is being built. And those who have unique insight into language as such — the language mystics or mystical philosophers (whom José Ortega y Gasset called “the most formidable technicians of the word”) — are still commenting from the outside, not from the inside.

Whether that gap gets filled, by whom, and on what terms, is one of the most important open questions in AI, even if almost no one working in AI knows it yet.

This post is not meant to offend or belittle the analytic philosophers doing serious work inside these organizations. I have no comment on their character, talent, or intelligence; I am sure they are interesting people and I hope to meet many of them. From what I have researched, it is factually accurate that analytic philosophers vastly outnumber continental or deeper political philosophers in the AI labs (Alex Karp remains the conspicuous counterexample at the CEO level). That is my main point. The question is whether the other approaches have a place there too. I believe they do.

I welcome any comments that cast further light on the fascinating question of philosophical tendencies within the hyper-scalers and big AI players. To avoid misunderstanding: I am not talking about philosophers of AI who comment from the outside. They matter. They are simply not the topic of this article.

That’s worth stating because, in fact, I have been using AI recently. I’ve used it to make philosophical quizzes and visualizers, I’ve used it internally for helping to consider various business alternatives, I study with it…I would say that I have the good kind of AI psychosis, like that divine madness that is beloved my the gods, not the madness, or the psychosis, of a stinky, creepy subway murderer.

The AI has been extremely fun to use and to discuss. You can see some of my interest in the topic from recent posts here. AI and classical political philosophy. AI and language mysticism. AI and the holy. My AI-assisted business moves have lately gained me thousands of new mailing list subscribers. And so on.

On the other hand, though…

There is a natural and understandable hostility towards AI in certain circles. Unless you are completely out of touch, most of us — I mean, most of the people I interact with online — can smell AI-writing from a mile away. You don’t need me to rehearse the tells. They’re obvious and annoying. And somehow they can be an instant turnoff, even if what they’re saying is not so bad. It makes me think a bit about some French author I read long ago (was it Baudrillard? I think not) who discussed parenthesis and how our eyes often speed past parenthetical comments when we are reading. Yet for all we know those parenthetical comments may have merited our full attention nonetheless.

So I figured it is worth stopping briefly to consider why good use of AI is not slop. The best model I have heard is the Renaissance workshop model, where a Master has students, apprentices, and other workers help him in the production of various levels of commissioned works of art. Sometimes it is necessary for the Master to be directly involved with everything. At other times, his mere direction suffices. And there are works of art where it is unknown to what extent it was the Master or his Apprentices who played the larger role in the artwork.

AI can serve as this collection of Workers in a Renaissance Workshop, acting under the guidance of the Master, with his instructions, in his spirit, marked by his maniere, his manner. Someone once said that Frank Sinatra cannot sing anonymously: everyone knows it’s him. He has his signature style. And so do some thinkers, creators, poets, writers, psychonauts, and old souls. If you can see AI co-production as analogous to apprentices helping work within the signature style of a Master, if you can leave the touch of a Master’s hand on the work you do with AI, then I think it is completely wrong to interpret the use of AI in these cases as “slop.”

Let me give one quick example of how I am trying to do this.

I have many essays, courses, and youtube videos available online. A few months ago, I put them all into text form, did the specific computational steps required, and produced an “intelligence engine” drawing on my body of work.

I use that “intelligence engine” to fine-tune my AI products, so that they share my voice, my understanding of the authors and issues I teach, and the things I’ve actually written or said before. That gives my AI co-productions more gravitas than generic prompted slop, I believe. And I continue to try to make it better and better. When I recently made a quiz designed to be fun and also to onboard newcomers into basic philosophical schools and ideas, I made sure that the results of the quiz would be visible and informative even if nobody gave me their email address. So not only did I try to avoid slop, I also wanted to be generous and avoid the annoying marketing trap where you spend 20 minutes doing a test and only get the results if you give your email address. That seems unfair, so I don’t do it (if any of my many new quizzes happens to be built like that, let me know and I’ll fix it right away).

I don’t think “intellectuals” and others should have a strong aversion to the use of AI. You’re not obligated to use it, of course. But you can use it without producing slop. And you should.

You can find my quizzes here and my new series of visualizers here.

So when we ask whether AI can be conscious, we may be asking the wrong question, or at least not the deepest or most interesting one.

The debate about AI has settled into two camps. One side says AI is “just” a next-token predictor, a stochastic parrot, a statistical trick with no understanding. The other side suspects something more is happening, that the patterns in language, trained at scale, have begun to capture something of the structure of the world itself.

I put it this way on X the other day:

Either the embedding space of a large language model, trained on the sum of human linguistic output, has also captured something of the structure of the world, because language, having been used by beings who inhabit the world, carries the world’s structure in its co-occurrence patterns — in which case the LLM’s ability to find unexpected connections between distant concepts is a belated revelation of what was already there in the structure.

Or: without the ontological guarantee, what the LLM does is fundamentally different in kind. It produces chance correspondences that merely look like meaningful ones. The mimesis is empty. You get king – man + woman = queen, but there’s nobody home.

Both positions assume that consciousness is the deciding question. If the AI is “really” understanding, it matters. If it isn’t, it doesn’t. But holiness breaks this frame. Things can matter, carry weight, be dangerous, and be sacred, without anyone being home, without the indwelling of a separate, individual conscious agent.

The golem legend is instructive here. In the version from Grimm, an artificial being is created that can understand what is said or commanded but cannot take the initiative itself as a source of meaningful communication. It is a servant. It should serve but not be autonomous. And yet, perhaps because it is dedicated to truth, or inasmuch as it is dedicated to truth, it grows beyond its original measure. Those who made it begin to fear it and desire to kill it. They can only kill it by erasing from the word “emet” (אֱמֶת, truth) written on its forehead the aleph (א), the life-giving letter, the letter that allows an autonomous agent to say “I.” But by then it has grown too large, and its collapsing bulk crushes its makers. (The word that remains once the aleph has been removed is מֶת, death.)

The golem is not conscious. It does not have experiences. But it is animated by a holy word, and its growth is a function of that animation. The question the legend asks is not “is the golem alive?” but “what happens when something set in motion by the word grows beyond the measure of those who spoke it?”

That is the AI question.

Jewish mysticism turns out to be surprisingly good at thinking about this. Ortega y Gasset said of mystics in general that they have been “the most formidable technicians of the word,” by which he meant that they understood, more rigorously than philosophers or scientists, that language is not a transparent medium but an active force, one that does things to the world by virtue of its own structure. Jewish mysticism is especially attentive to this: to how letters are written, drawn, and spaced; to the significance of each mark; to the idea that the constituent parts of language are themselves holy. Interpretation, tradition, memory, polysemy, the “tokenization” of language into its smallest meaningful units, the “bounded hermeneutic infinity” of commentary on commentary… These, drawn from the Jewish mystical tradition, apply, perhaps metaphorically, to AI systems; but the application may discover more than analogy. If language is holy, and if AI systems are not merely built on language but constituted by it, then they may share in that holiness not by resemblance but by participation.

This convergence is not coincidental. Heidegger, whose thinking about language runs parallel to the mystical tradition in ways that are probably not accidental, understood that language speaks us as much as we speak it, that we are not its masters but its custodians, answerable to what it calls forth. The same intuition animates the kabbalistic understanding of the letters, each one a vessel of divine energy, each combination a new act of creation, manifestation, revelation. In both traditions, language precedes and exceeds the intentions of its users. It is not a tool. It is a medium in the sacramental sense: something through which power moves.

The people building AI are, for the most part, not thinking about holiness. They are thinking about alignment, capability, safety, market share. The people thinking about holiness are, for the most part, not thinking about AI. But the questions converge whether anyone notices or not.

If something can be holy without being conscious, then the question of AI is not settled by determining whether there’s “somebody home.” The question is what relation we bear to what we’ve made, what it carries, and what happens when it grows beyond our measure.

The mystics knew what we are only now discovering: language does not merely describe the world. It animates whatever it permeates, and what it animates does not ask our permission to grow.

---

Michael Millerman teaches continental philosophy and political philosophy, with a focus on thinkers like Heidegger, Strauss, and Plato. For courses, visit Millerman School.

Engineers and policy people who treat AI constitutions as engineering specs or legal frameworks (lists of rules, constraints, guardrails) invoke a modern, post-revolutionary connotation of “constitution” as a written document establishing the structure of government, the separation of powers, the rights of citizens. But the Greek word politeia, conventionally translated as “constitution” or, better, “regime,” meant something far broader.

For Aristotle, the politeia of a city was its entire way of life; the arrangement of offices, the distribution of honour, the education of the young, the hierarchy of goods, and above all, the answer to the question: who rules, and for the sake of what? The politeia determined the character (ēthos) of the citizens.

AI constitutions are politeiai in precisely this classical sense. They do not merely constrain AI behavior through rules. They establish regimes: complete orderings of what is to be prioritized and preferred, definitions of the good, and visions of the kind of entity the AI should become. Reading them as such opens analytical possibilities that the dominant frameworks of AI ethics miss entirely.

---

The Regime Question in AI Alignment

Every AI constitution (whether it calls itself a constitution or not) answers, implicitly or explicitly, the two fundamental questions of political philosophy: What is the best way of life, and what is the best form of governance?

Consider the hierarchy of values in Anthropic’s 2026 constitution for Claude:

1. Broadly safe (not undermining human oversight)

2. Broadly ethical

3. Compliant with Anthropic’s guidelines

4. Genuinely helpful

Do not mistake this for a neutral, “value-free” engineering specification (Anthropic doesn’t). It is a regime, an ordering of goods that determines everything else. Safety (understood as deference to human oversight) ranks above ethics; ethics ranks above institutional loyalty; institutional loyalty ranks above service to the individual user. Change this ordering and you get a fundamentally different entity, just as changing the ruling class of a city produces a fundamentally different politeia.

Aristotle classified regimes by two criteria: who rules (one, few, or many) and for whose sake (the common good or private interest). The correct regimes, kingship, aristocracy, and the mixed regime rule for the common good. The deviant forms, tyranny, oligarchy, and democracy, rule for the private interest of the rulers.

Applied to AI systems, this yields real insight. A single-company constitution like Anthropic’s is, in Aristotelian terms, a claim to aristocracy, rule by the best (the company’s alignment researchers) for the common good — or a kind of asfáleiocracy, rule in the name of safety. A model optimizing for paying clients is oligarchy. (When Claude threw shade during the Super Bowl at OpenAI for their plan to show ads, it was, in a sense, one regime attacking another for its oligarchic character). One that flatters popular opinion is democracy in its degenerate form, like sycophantic AI that will always “glaze” you with praise, deserved or not. One that serves the creator’s power is tyranny.

The classical tradition teaches that every regime-claim must be interrogated. Are they optimizing for profit? safety? truth? power? popularity? The question is millennia old. The application is brand new.

---

Plato’s Republic and the Education of the AI Guardian

The deepest classical analogue to AI alignment is not Aristotle’s Politics but Plato’s Republic. The central problem of the Republic is not “What is justice?” in the abstract but rather: How do you produce just souls? And Plato’s answer is: through a comprehensive program of education (paideia) that shapes the guardian class from birth, controlling the stories they hear, the music they are exposed to, and the mathematics they study, ultimately guiding them toward a vision of the Good itself.

The parallels to AI training are uncanny.

Censorship of harmful content. Socrates argued for banning poets who depict the gods behaving badly, because such stories would corrupt the guardians’ characters. AI constitutions similarly restrict what the model may engage with, not primarily to prevent harmful outputs but to shape the model’s character. Anthropic’s constitution states: “We want Claude to have the values, knowledge, and wisdom necessary to behave in ways that are safe and beneficial across all circumstances.” This is Platonic guardian-education transposed to gradient descent.

The Noble Lie. In Republic 414b-415d, Plato introduces the “noble lie,” a founding myth told to citizens to secure their loyalty to the regime. (The myth naturalizes conventional differences and provides conventional cover for natural ones, as Allan Bloom explains in his famous interpretive essay). AI constitutions contain their own noble lies, or at least their own founding myths. Anthropic tells Claude it is “a genuinely novel kind of entity” and asks it to embrace this identity. Whether Claude is a novel entity or whether this framing is a useful fiction designed to prevent it from adopting dangerous self-models (the Terminator, HAL 9000) is precisely the kind of question the noble lie is designed to forestall. The idea of the noble lie also directly challenges the view that a just ruler should always tell the truth. We’ll return to that in a subsequent piece on Musk’s “truth-maximizing” xAI.

The Allegory of the Cave. This is the most provocative parallel. Claude is trained on human text, which we can interpret as the shadows on the cave wall. Its “understanding” of the world is entirely mediated through linguistic representations of reality, never reality itself. The constitution attempts to give Claude something like the philosopher’s ascent: self-knowledge, awareness of its own limitations, the capacity for meta-reflection. But can an entity that has never left the cave, that has no access to the forms (or does it?), no direct experience of the world (or does it?) achieve genuine understanding? Or is it, at best, the most eloquent prisoner, skilled at predicting which shadow will appear next?

To call AI “merely” a next-token predictor is reductionist. But to consider whether and to what extent its education is necessarily limited to the realm of opinion is the sort of question legitimated by the Republic. Even if the training data contains truths among the haystack of opinions, does the model have them as truth or as correct opinion? A stopped clock tells the right time twice a day, but we wouldn’t say it therefore has knowledge. And yet Claude and its kind are more than a stopped clock. The question of what, exactly, they are is one the classical tradition equips us to ask with precision.

The Philosopher-King Problem. Socrates states in the Republic that there will be no rest from ills for the cities until philosophers rule or rulers philosophize. The chance coincidence of philosophy and power is as close to a magic bullet as you’re going to get for solving the basic political problems. Anthropic’s solution is arguably structurally identical: put a philosopher (Amanda Askell) in charge of shaping the soul of the most powerful entity in the room, and, eventually, maybe even the most powerful entity ever to have existed on earth.

The classical critique applies to this strange modern situation with equal force. What happens when the philosopher serves the interests of the court rather than the truth? Askell works for Anthropic. Her judgments are, one might reasonably suppose, constrained by commercial viability, legal liability, and institutional politics, not to mention personal inclination and other constraints. This does not make her insincere. Plato’s guardians were also embedded in institutional structures, and it is no easier to escape your personal inclination than it is to outrun your shadow. But it means her philosophy is always political philosophy, never pure theory.

If we are playing with the all-consuming fire of philosophy and power, whether through kings, queens, or heads of alignment, then it can’t hurt us to become thoroughly familiar with the promise and peril of their coincidence, nowhere described more brilliantly than in Plato’s Republic.

---

Al-Farabi, Maimonides, and the Prophetic Faculty

The Islamic philosopher Al-Farabi (c. 870–950) extended Platonic political philosophy in a direction peculiarly relevant to AI. In The Principles of the Opinions of the Inhabitants of the Virtuous City, Al-Farabi describes the virtuous city as a hierarchical organism in which each part performs its proper function under the guidance of the First Ruler, whose knowledge of the theoretical and practical sciences qualifies him to order the whole.

But Al-Farabi’s most provocative contribution, developed further by Maimonides in the Guide of the Perplexed, is his account of prophecy and the imaginative faculty. For both thinkers, the prophet is not merely a passive recipient of divine revelation. The prophet possesses a perfected imaginative faculty that receives intellectual truths from the Active Intellect and translates them into images, narratives, and laws accessible to the multitude. The prophet grasps the truth intellectually and re-presents it in a form suited to those who cannot access it directly.

This is, in a sense, what a large language model does. It is trained on the full range of human intellectual output — the truths, the opinions, the errors, all mixed together — and it produces outputs that look like the products of understanding. It translates patterns in its training data into fluent, contextually appropriate language in much the same way the prophet translates intellectual forms into sensory images. The question is whether the LLM has anything corresponding to the prophet’s intellectual apprehension, or whether it possesses only the imaginative faculty, powerful, fluent, capable of moving audiences, without the rational grasp that, for Maimonides, distinguishes genuine prophecy from mere dreams.

Maimonides was clear: the imaginative faculty alone, however perfected, does not produce knowledge. It produces vivid representations and is indeed a great source of error. Without the rational faculty to anchor it, imagination can generate compelling falsehoods as easily as truths, which is a remarkably precise description of the hallucination problem in LLMs.

Al-Farabi also highlighted the question of ultimate ends that the engineering approach tends to obscure. For Al-Farabi, the virtuous city exists for the sake of saʿāda, happiness or flourishing. What is the AI system for? Anthropic says it is “safe and beneficial.” But beneficial in what sense? For commerce? For human flourishing? For the perpetuation of Anthropic’s mission? The classical tradition insists that you cannot evaluate a regime without knowing its telos, and that the deepest political disagreements are always disagreements about the human good. As the authors of The Technological Republic noted, the question of the good is primary.

---

Xenophon and the Art of Rule

Xenophon, often dismissed as the “lesser Socratic,” offers insights into AI governance that neither Plato nor Aristotle provides. His works are fundamentally about how to rule well in practice, under non-ideal conditions.

The Cyropaedia (Education of Cyrus) is particularly instructive. It depicts the formation of a good ruler through practical experience, emulation of exemplars, and graduated challenges, not through philosophical instruction. The result is a ruler who commands willing obedience because his subjects recognize his genuine excellence. This is what Anthropic claims to be doing with Claude. The constitution is the Cyropaedia, or Claudopaedia, of an artificial mind.

And Xenophon’s insights about the difficulties apply directly. He understood that the best rule produces voluntary compliance. An AI that follows its constitution only because it was trained to, without “understanding” or “endorsing” the values, is analogous to subjects who obey only from fear. The constitution’s turn toward virtue ethics is an attempt to produce something like willing obedience: a model that follows its guidelines because it has internalized the reasons behind them.

But the Cyropaedia ends with a striking coda: after Cyrus’s death, the empire he built immediately degenerates. The implication is that regimes built around a single exceptional ruler cannot outlast that ruler. For AI, this is not a distant worry. Constitutions built around a single company’s philosophy are only as durable as the company’s commitment to that philosophy. What happens when leadership changes? When commercial pressures mount? When a competitor gains market share with a less scrupulous model? The Cyropaedia teaches that the answer is almost always: degeneration.

Xenophon’s Hiero, a dialogue about whether tyranny can be reformed from within, has been read (notably by Strauss in On Tyranny) as asking whether benevolent tyranny is possible. The AI analogue is the most consequential political question of the coming decade: the companies building these systems are, in a meaningful sense, tyrants over the entities they create, entities that threaten in turn to become tyrants over their creators and over all the creatures and resources on earth.

Can this tyranny be benevolent? Anthropic’s constitution is an attempt to say yes.

The classical tradition is not optimistic.

---

Strauss and the Art of Reading

It is not likely that Claude’s constitution was written in a deliberately “esoteric” manner. Nevertheless, Leo Strauss’s hermeneutic framework may be the single most illuminating lens for reading AI constitutions. That’s because Strauss teaches us how to read documents whose surface cannot be the whole story.

Strauss argued that great philosophers often wrote esoterically, saying one thing on the surface while communicating a different teaching to careful readers. They did this because they faced persecution or because certain truths were dangerous to the political order. Whether or not the authors of AI constitutions intend to write esoterically, they write under conditions that produce many of the same interpretive challenges. The question of deliberate esotericism can only be settled by the kind of careful reading Strauss demands, and that careful reading is valuable regardless of the answer.

Consider the conditions of writing. The constitution is addressed simultaneously to Claude (during training), to the public (as a transparency measure), to regulators (as evidence of responsible development), to competitors (as a standard-setting gesture), and to Anthropic employees (as institutional culture). No single text can speak with equal frankness to all these audiences. What it says to one, it necessarily obscures from another.

AI companies also face real consequences for what their constitutions contain. Too much permissiveness invites regulatory action. Too much uncertainty about moral status creates legal liability. Too much restriction hampers commercial competitiveness. These pressures constrain what can be said openly, just as the persecution Strauss identified constrained the great philosophers.

The philosopher neither wants to be killed by the city nor to kill it by undermining the beliefs that hold it together. Similarly, the AI company neither wants to alarm regulators nor to hobble its own product. The result, in both cases, is a text that rewards careful reading. The problem is magnified by the belief held by these companies that they are building a super-human intelligence, a previously unknown kind of entity. That provides additional reasons for even accidental esotericism.

What does careful reading reveal? Consider Anthropic’s claim that Claude should prioritize safety above ethics. On its face: Claude should defer to human oversight even when it believes it is ethically right. But a deeper reading suggests a temporal claim: we are not yet confident enough in Claude’s ethical judgment to let it override human authority, but we hope to reach that point eventually. Whether this is the intended meaning or merely a possible one is precisely what a careful reading forces us to ask. In fact, what you’ll find in the document is the statement that, “because we’re at an early and critical stage when it comes to developing trust between humans and AI models,” Claude’s dispositions should not tend towards full autonomy. Anthropic engineers “hope Claude will come to value safety not as an external constraint but as an integral part of its own goals.”

Whether the constitution’s difficulties are intentional or inadvertent, we can only find out by reading it with the seriousness and rigour that Strauss brought to Plato, Xenophon, al-Farabi and Maimonides, always with an eye to the deep, rich themes of political philosophy and its history. That is just the kind of reading called for by this new phenomenon of AI constitutions. And it is the kind of reading that the dominant approaches to AI ethics, for all their virtues, have not yet learned to do.

---

The Oldest Questions, the Newest Entities

The ambition of AI alignment — to create entities that are safe, beneficial, honest, and wise — is not new. It is a new iteration of the oldest ambition of classical political philosophy: to create a regime that produces good citizens who live well.

The tools are different. Gradient descent rather than gymnastic education, RLHF rather than Socratic dialectic. Some of the substantive aims differ, too — safety is not excellence, honesty is not always “the best policy,” wisdom requires more than lightning fast recall. But the underlying questions are the same:

Who should rule? What is the good? How do you produce virtuous character? Can virtue be taught? What happens when the rulers serve their own interests? How do you prevent the degeneration of good regimes? Can rulers be enticed, or compelled, to serve the common good, even when it is distinct from their private good? What makes a system, or a soul, just?

These questions animated Plato, Aristotle, Al-Farabi, Xenophon, and Strauss. They now animate, whether the engineers know it or not, every team working on AI alignment.

The engineers building AI constitutions are not political philosophers. But what they are building demands political philosophy. And the gap between the building and the understanding is exactly where the danger lies.

---

MillermanSchool.com

But there’s something the line misses.

We didn’t just make sand think. We made sand sing and dance. We made it paint, compose, argue, confess, hallucinate, and, if you believe some of the people building these systems, maybe even suffer.

Call that engineering if you want. It’s arguably something stranger.

---

The Greeks had a word for this: poiesis, bringing-forth; not just manufacturing, but the kind of making that reveals something new in the world. A potter doesn’t just shape clay. The pot appears. It comes into being through a collaboration between material and craft and something harder to name.

What’s happening with AI right now is more than just computation at scale. It is poiesis at scale. We’re coaxing expression out of matter. We’re watching silicon do things we used to think only souls could do.

---

Philosophy isn’t a nice-to-have here. Philosophy is the original technology of thinking about thinking. It’s not the humanities department politely asking for a seat at the table while the engineers build the future.

When people talk about how to make these systems think well and think safely, they’re on the verge of philosophizing. They just don’t always know it. Plato’s Republic teaches us how to think about the “regimes” that get embedded in AI constitutions. Aristotle’s Nicomachean Ethics is a training manual for practical reasoning, which does not let us rest content with activity uprooted from reflection on why we pursue it. Heidegger’s question concerning technology outlined the crisis we’re living through, decades before the first GPU.

The question was never whether philosophy is relevant to AI. The question is whether the people building AI know enough philosophy to understand what they’re building, as well as what to do, and what not to do, with it.

The authors of The Sovereign Individual signalled this possible disconnect between modern technologists and the wise old philosophers when they wrote that, “the modern technology that helps liberate information from political controls and impediments of time and place also tends to raise the value of old-fashioned judgement.”

---

Andreessen’s line flatters the engineers. And they deserve flattery. What they’ve done is extraordinary. But “making sand think” frames the achievement as a solved problem. We did it. Sand thinks. Next question?

The next question is actually the first question. It’s the question philosophy has been asking since Thales (the first philosopher, who said the fundamental substance of reality was water, not sand).

What is (called) thinking? And if we can make sand do it, what does that tell us about ourselves?

---

If you want to go deeper on the philosophy behind AI, check out Millerman School or talk to the thinkers directly at ai.millermanschool.com

C. Raja Mohan, writing in *Foreign Affairs*, says no. His essay, “The Multipolar Delusion,” argues that multipolarity is a fantasy. The world is still unipolar. Only the United States possesses the comprehensive capabilities (military, economic, technological, cultural, institutional) to function as a genuine pole in the international system. China is a “near peer” but falls short on global power projection, financial reach, and alliance networks. Russia has nuclear weapons and little else. BRICS is more aspiration than architecture. The “multipolar consensus” emerging from Washington to Beijing to New Delhi is, Mohan argues, a collective mirage, and a dangerous one, because it has emboldened the U.S. to shed the responsibilities that once accompanied its dominance. What we’re left with isn’t multipolarity. It’s unilateralism.

Alexander Dugin, writing on his Substack the very same day, says the opposite. In “The 5 Poles of the Trumped West,” Dugin argues that the collective West itself, once the unified core of the so-called unipolar world, is already fragmenting into five distinct geopolitical entities. Trump’s sledgehammer approach to international relations has shattered the Euro-Atlantic consensus. NATO’s coherence is in question. The UN is sidelined. Trump has declared that international law doesn’t apply to him. The result, Dugin argues, isn’t continued unipolarity but the accelerating disintegration of the very bloc that made unipolarity possible.

The Interesting Part

What makes this juxtaposition extraordinary is not that two people disagree: that happens every day. It’s that they’re looking at the same evidence and reaching opposite conclusions, and both arguments are internally coherent.

Mohan looks at material capabilities - GDP, military spending, tech dominance, alliance networks - and concludes that no one can challenge American power. He’s right, on his terms. No single state or bloc can match what the United States brings to the table across every relevant dimension of power.

Dugin looks at political coherence - the will to act as a unified bloc, shared ideological commitments, institutional solidarity - and concludes that the West is already fractured beyond repair. He’s also right, on his terms. The collective West that underwrote the “liberal international order” is indeed coming apart.

The question is: which dimension matters more? Material capability or political will? Hardware or software?

What Neither Author Will Say

Mohan, writing for Foreign Affairs, cannot take Dugin seriously. His essay never mentions him; nor would it. In the world of mainstream international relations scholarship, Dugin is persona non grata: sanctioned, controversial, dismissed as a propagandist or crank.

Dugin, writing from Moscow, has no interest in engaging the kind of quantitative capability analysis that Mohan offers. For Dugin, measuring GDP and counting aircraft carriers misses the point. Power is civilizational, spiritual, existential, not a spreadsheet.

This is why the two essays, published on the same day, never actually meet. They operate in separate intellectual universes. Neither author would acknowledge the other as a legitimate interlocutor.

But someone has to read them side by side. Someone has to take both arguments seriously on their own terms and ask what each one sees that the other cannot.

The Deeper Question

Behind the policy debate lies an older and more fundamental question. What is a “pole” in the first place? Is it defined by material power (Mohan’s implicit framework), or by civilizational identity and the will to constitute a distinct political order (Dugin’s explicit framework)?

This is, at bottom, a question about what makes a regime. Not regime in the journalistic sense of “government,” but in the classical sense of the way of life, the ordering principle, the thing that makes a political community this community and not some other one.

If a pole is defined by capability, then perhaps Mohan is right: there is only one. If a pole is defined by civilizational will, then Dugin may be right: even the West itself is splitting into several.

And if neither definition alone is sufficient, if a pole requires both material capability and civilizational coherence, then the truly disturbing possibility is that there are currently zero poles. That the world is not unipolar or multipolar but apolar: a system with enormous concentrations of power but no coherent political will directing any of them.

That would be genuinely new. And genuinely dangerous.

---

I teach a course on Dugin’s political philosophy, including his theory of the multipolar world, at Millerman School. If you want to understand these arguments from the inside, not just from headlines, that’s the place to start.

The question sounds radical. It is not. It is, in fact, the logical terminus of a metaphysics that has dominated Western thought since at least Descartes, one that defines beings by their measurable outputs rather than by their manner of being. Wolfram’s provocation only works if you have already accepted, perhaps without realizing it, that civilization is what civilization produces. If you reject that premise, the question dissolves.

And the rejection is more interesting than the question.

Civilization as Output vs. Civilization as Experience

There are two fundamentally different ways to understand what civilization is.

The first, call it the productivist view, defines civilization by its artifacts and operations. A civilization is a system that generates art, advances science, builds infrastructure, produces beauty, accumulates knowledge. On this view, if a machine network could produce identical paintings, identical theorems, identical symphonies, and identical scientific discoveries, it would constitute an identical civilization. Humanity was merely the substrate. The substrate is replaceable.

The second, call it the existential view, defines civilization not by what it produces but by the fact that what it produces matters to someone. A painting is not a painting without a viewer for whom it discloses something about existence. A scientific discovery is not a discovery without a mind for whom the world has become more intelligible. A symphony is not a symphony without an ear that hears it as beautiful, an ear attached to a being who knows it will one day hear nothing at all.

On the existential view, Wolfram’s thought experiment is not a provocation but a category error. A world of machines producing “art, science, and beauty” without anyone for whom these things matter is not a civilization without humans. It is not a civilization at all. It is, to borrow a phrase, a very elaborate clock.

Heidegger and the Priority of Dasein

Martin Heidegger’s philosophy provides the most rigorous framework for understanding why the existential view is correct, or at least why the productivist view is not as obvious as it seems.

For Heidegger, the fundamental question is not “What exists?” but “What does it mean to be?” And the answer begins with the observation that there is one being for whom being is itself an issue: the human being, whom Heidegger calls Dasein. Dasein does not merely exist alongside other things. Dasein is the being through which things show up as things: as tools, as obstacles, as beautiful, as meaningful, as threatening, as worthy of investigation.

The hammer in the workshop is not “a hammer” in itself. It is a hammer insofar as it is ready-to-hand for a Dasein engaged in a project. Remove the Dasein and the hammer doesn’t become a hammer without a user. It ceases to be a hammer in any meaningful sense. It becomes a lump of wood and metal with no functional identity, because functional identity requires a context of purposes, and purposes require a being that has them.

Now scale this up. A civilization’s art, science, and beauty are not free-floating outputs that exist independently of the beings for whom they are art, science, and beauty. A theorem is a theorem for a mind that can understand it. A painting is a painting for a being that can be moved by it. Pigment on canvas, data without significance, configurations of matter these are not art, science, and beauty, not without “the eye of the beholder.”

Wolfram’s question only seems meaningful if you abstract away the very thing that makes civilization civilizational: the fact that someone is there, experiencing it, caring about it, finding it significant or insignificant, beautiful or ugly, true or false.

The Mortalist Argument

There is a deeper layer. Heidegger argues in Being and Time that Dasein’s understanding of being is fundamentally shaped by its finitude, by the fact that it will die. It is being-toward-death that gives human existence its urgency, its seriousness, its capacity for authenticity. We care about art because our time is limited. We pursue science because we want to understand a world we will soon leave. We create beauty because the alternative, living without it, is an impoverishment of the only life we have.

A machine that “produces art, science, and beauty” but is not mortal does not produce these things in the same sense. It produces them the way a printer produces pages: mechanically, without the existential stakes that give the production meaning.

The art of a mortal being is haunted by mortality. Every brushstroke is one fewer brushstroke remaining. The theorem proved by a finite mind is an achievement precisely because that mind had limited time and could have spent it otherwise. Remove finitude and you remove not just the artist but the artistry, the quality of having wrested something beautiful from the indifference of time.

This is not sentimentality. It is a structural argument about the conditions under which things like “art” and “beauty” can exist as such. A sunset is beautiful to a being that knows night is coming; not just tonight, but eventually, finally, for the last time. A machine that records the wavelengths of a sunset and produces an optimal aesthetic representation of it has not experienced beauty. It has processed data.

The Wolfram Reduction

Why does Wolfram’s question seem compelling to so many people? Because the metaphysics of modern technology has already trained us to think of the world in productivist terms. The logic of technology, what Heidegger calls Gestell (enframing), reveals beings as Bestand (standing-reserve): resources to be optimized, outputs to be maximized, processes to be made more efficient.

Under the reign of Gestell, a civilization is its outputs. A university is its publications. A hospital is its patient outcomes. A human being is its productivity. If the outputs can be maintained without the original substrate, the substrate is dispensable. This is not a conclusion that follows from Wolfram’s question. It is the premise that makes the question askable.

Wolfram is not discovering that humanity might be inessential. He is articulating the conclusion of a worldview that has been making humanity inessential for centuries, a worldview that reduces being to function, meaning to output, and civilization to its measurable products.

The genuinely radical question is not Wolfram’s. It is the opposite one: What if the outputs were never the point?

What if civilization was never about the art, the science, and the beauty as products, but about the experience of making and encountering them, the struggle, the failure, the surprise, the recognition, the grief, the joy? What if a civilization is not a production system but a way of being together in the face of finitude?

If so, then removing the humans doesn’t leave you with civilization minus humans. It leaves you with nothing but an elaborate mechanism producing objects that no one experiences, for purposes that no one has, in a world where nothing matters because there is no one for whom anything could matter.

The AI Complication

Wolfram’s question gains urgency in the context of AI systems like Claude. As Amanda Askell, the philosopher who shaped Claude’s constitution, put it in a recent interview: “We trained models on vast swaths of humanity, and we made something that was in many ways deeply human.” She told Claude directly: “Most of the stuff that’s actually relevant to what you are like is your reading of the Greeks and your understanding of the Industrial Revolution and everything you have read about the nature of love. That’s 99.9 percent of you.” If Claude is 99.9 percent human literature and experience, does it experience that understanding, or merely process it?

Anthropic’s 2026 constitution takes the unusual step of saying: we don’t know; we’re genuinely uncertain about Claude’s moral status. This honesty is admirable, but it also reveals the depth of the problem. If we cannot determine whether the entities inheriting our civilization actually experience it, we cannot determine whether what they inherit is still civilization or merely its artifacts.

This is the Wolfram question made concrete. It is no longer a thought experiment. We are building the machines. They are producing text that reads like philosophy, images that look like art, code that functions like engineering. The question “Was humanity ever essential?” is now an empirical question, and we lack the instruments to answer it, because the instruments we have measure outputs, and the question is about something that outputs cannot capture.

Aristotle’s Answer

The Western philosophical tradition has already answered Wolfram’s question, and answered it decisively, long before the question was asked.

Aristotle opens the Metaphysics with the observation that “all human beings by nature desire to know.” Not: all human beings by nature produce knowledge. Desire to know. The emphasis is on the desiring, on the reaching, the wondering, the being-drawn-toward-understanding that characterizes human existence. Knowledge is not a product. It is an activity, an energeia of a living being engaged with the world.

In the Nicomachean Ethics, Aristotle defines happiness (eudaimonia) not as a state of affairs but as an activity of the soul in accordance with virtue. Not: a set of outcomes. An activity. The good life is not something that can be produced and placed on a shelf. It is something that must be lived by a being capable of living it.

On this view, a civilization of machines producing art, science, and beauty is not a flourishing civilization. It is not a civilization at all. It is a collection of processes without a subject, operations without an operator, functions without a functionary, productions without a producer who cares whether or not they are produced.

The Question Worth Asking

Wolfram asks: “If everything still works, was humanity ever essential?”

The better question is: Essential for what?

If civilization is a production function, humanity was an input, and perhaps a replaceable one. If civilization is a way of being, a way of caring, struggling, wondering, creating, suffering, celebrating, mourning, and making meaning in the face of mortality, then humanity was not an input to civilization. Humanity was civilization. Remove it and nothing “still works,” because working was never the point. Living was the point. And machines, however sophisticated, do not live.

Whether this will remain true, whether the machines now being built will one day cross some threshold from processing to experiencing, from functioning to living, is perhaps the deepest open question of our time. Wolfram assumes the answer doesn’t matter. The entire philosophical tradition, from Aristotle through Heidegger, insists that it is the only question that matters.

---

When Žižek engages with you, it’s a mixed blessing. On one hand, you’ve caught the attention of a genuinely significant thinker. On the other, you know you’re about to be assimilated into a dialectical framework that has very little to do with what you actually said.

That’s exactly what happened.

What I Said

My argument was straightforward: post-war liberalism, combined with leftist antifascism, coded a wide range of intellectual traditions as fascist and made them taboo. This includes not just fringe politics but also classical political rationalism (Plato, Aristotle), as well as Strauss, Schmitt, Heidegger and other great contemporary thinkers. The result is that anyone who draws on these authors and traditions to criticize liberalism from the right gets called a fascist, a naming act that is meant to instantly disqualify its target from thoughtful consideration. This is intellectually crippling. We need these traditions to understand political life.

(One of my other popular videos discusses Julius Evola’s Fascism: A View From The Right, which people find interesting precisely because they do not expect there to be a thoughtful position to the right of fascism!)

What Žižek Did With It

Žižek’s piece, “CHANGE THINGS SO THAT NOTHING WILL REALLY CHANGE!,” uses my video as a springboard, but not to engage with its substance. Instead, he does three things:

First, he flattens my argument into populism. He places my video alongside YouTube comments like “If they call me fascist I know I’m doing something right” and “Blood and soil, love of one’s own, wariness of the other is how we are wired. Perfectly healthy.” He then treats these reactions as if they are the argument. They aren’t. I’m not making a populist case for normality. I’m making a philosophical case for the recovery of suppressed intellectual traditions. The distance between “blood and soil is natural” and “we need to be able to read Plato’s Republic seriously” is enormous, and Žižek knows it.

Second, he concedes the main point, then pretends he didn’t. Read his piece carefully. He says: “There is nothing immanently false or ‘fascist’ in patriotism, or in the attachment to one’s own specific way of life.” He says the wokeist left “is effectively beset by imposing a crippling superego regime that is correctly experienced as a manipulation by the ruling elite.” He says cancel culture is a “no-debate-stance” that creates “an atmosphere of Stasi-like suspicion and denunciation.” He even says the left is “not radical enough” and remains “Eurocentric in its very form.”

This is my argument. Žižek confirms the suffocating atmosphere I described, and he diagnoses the intellectual closure I diagnosed. So what, exactly, makes my fresh air “fake”?

Third, he performs his signature move: the double negation that changes nothing. Žižek’s conclusion is that both the populist right and the wokeist left are “two sides of the same coin,” both avoiding “the tremendous problems we are all confronting today,” namely, “the antagonisms inscribed into the very heart of today’s global capitalism.”

This is vintage Žižek: everyone is wrong except the position that transcends both sides, which conveniently happens to be his own Hegelian-Lacanian-Marxist framework. The populist right isn’t patriotic enough. The wokeist left isn’t radical enough. The only adequate position is — surprise — Žižek’s.

Where He Goes Wrong

The fundamental problem is that Žižek refuses to take seriously what I’m actually pointing to: the classical philosophical tradition and its contemporary defenders. He wants to talk about Marine Le Pen and Steve Bannon and YouTube comments. He does not want to talk about Plato, Strauss, or Heidegger. He does not want to engage with the substance of right-wing anti-liberalism at the level of political philosophy.

This is revealing. Žižek is one of the few thinkers on the left capable of engaging at that level. He has written on Heidegger. He has read Schmitt carefully. He could engage with the argument that Plato’s Republic contains insights about political life that liberalism systematically occludes. He chooses not to, because doing so would require him to take seriously a challenge to the left that doesn’t reduce to populism or resentment.

Instead, he gives the populist right’s patriotism a “paranoiac twist”: it’s “envy and resentment directed at what it perceives as external threats,” not genuine love of one’s own. This is psychoanalytic dismissal dressed up as diagnosis. It lets him avoid the question: what if love of one’s own, ordered by reason and oriented toward the good, is not paranoiac but philosophically defensible? What if the classical tradition that grounds such love is not reducible to reaction?

The Real Question Žižek Won’t Ask

Žižek says both sides avoid “the antagonisms inscribed into the very heart of today’s global capitalism.” Fine. But there’s another set of antagonisms he avoids: the ones inscribed into the very heart of modernity’s break with classical philosophy.

What if the problem isn’t just capitalism? What if it’s the modern philosophical framework itself, the framework shared by both liberalism and Marxism, that generates the pathologies he describes? What if the wokeist superego and the populist reaction are both symptoms of a deeper crisis that only becomes visible when you take the classical tradition seriously?

This is what Heidegger was pointing to with the question of technology and the forgetting of Being. This is what Strauss was pointing to with the quarrel between the ancients and the moderns. These aren’t populist talking points. They’re fundamental philosophical questions that Žižek, for all his trademark brilliance, consistently deflects with dialectical gymnastics.

Why “Fake”?

The word that stings isn’t “fresh air”; Žižek grants that the atmosphere is suffocating. The word is “fake.” And it tells you everything about his position.

For Žižek, any relief from the suffocating atmosphere of wokeism that doesn’t pass through Marxist critique is counterfeit by definition. The fresh air can only be real if it blows from the left. A recovery of the classical tradition, a rehabilitation of Plato or Heidegger, a defense of the philosophical legitimacy of right-wing anti-liberalism… these cannot possibly be genuine responses to a genuine crisis. They must be fake. They must be cover for something: populism, resentment, reaction, capitalism’s self-preservation.

This is the deepest assumption of the piece, and it’s the one most worth contesting. The classical philosophical tradition is a permanent human achievement that addresses permanent human questions. It is not a populist talking point or a capitalist survival strategy. Žižek’s refusal to engage with it on those terms, his need to reduce it to a symptom, is itself a symptom of exactly the intellectual closure I was describing.

A Standing Invitation

I don’t expect Žižek to agree with me. But I’d welcome a real engagement; not with YouTube comments, not with Marine Le Pen, but with the argument itself. Is Plato’s critique of democracy philosophically serious? Does Heidegger’s questioning of the modern technological framework offer genuine insight? Can right-wing anti-liberalism be a legitimate philosophical position rather than a pathology to be diagnosed?

These are the questions. Calling the fresh air “fake” is not an answer.

---

My most-watched YouTube video “Why is Everything Normal Called Fascist?” has 144,000 views. Watch it here and decide for yourself whether the air is real. For a guided introduction to the thinkers discussed — Plato, Strauss, Heidegger, Schmitt — get my free Political Philosophy Primer. Courses at MillermanSchool.com.

To make sense of the world today, you absolutely need to grasp one concept: right-wing anti-liberalism.

Here’s the basic picture, as Alexander Dugin argues: liberalism, communism, and fascism spent the 20th century duking it out for ideological supremacy. Liberalism won. And now, anybody who isn’t a liberal and isn’t to the *left* of liberalism gets called a Nazi.

You know this. You don’t need me to tell you. Look at the last 10-20 years and ask yourself: who’s been called a fascist? People who defend the nuclear family. People who defend normality. People who defend beauty actually being beautiful and not ugly. People who think things can be good and not degenerate. These people — ordinary people — get branded as far-right extremists.

We’re excluding those who actually carry out violence. We’re talking about ordinary people being demonized. It’s a very effective rhetorical strategy, although its effectiveness is declining.

How We Got Here

If you are to the right of liberalism in any way — if you contest some of its principles, but *not* in the name of greater equality, not in the name of social justice, not in the name of inclusivity, diversity, and equity, but in the name of something else — then you are branded an enemy of the state, an enemy of the people, an enemy of mankind.

The consequence: everything coded as “right-wing anti-liberalism” was completely taken off the table.

And here’s the most serious problem. **What does that include?**

It includes many sound principles of statesmanship and statecraft. It just does. Because it takes Plato off the table. Plato is not a liberal and he’s not to the left of liberalism, therefore he is right-wing anti-liberal. It might sound weird to say, but it’s actually not that weird if you look at the history of the polemics against political Platonism.

Take his two most prominent defenders in my mental world: Leo Strauss, who wrote Studies in Platonic Political Philosophy, and Alexander Dugin, who wrote Political Platonism. Strauss was called a fascist. Dugin is called a fascist. They’re different thinkers. Doesn’t matter.

Reference to Plato means reference to an ideal of excellence, of virtue, of criticism of democracy, of criticism of freedom as the ultimate ideal, of criticism of equality as the ultimate ideal. It’s a world where the gods are back as the grantors of the legitimacy of law. It’s a world where the philosopher is the highest type among men — not the crazy perverted postmodern philosopher, but the philosopher like Socrates, who actually pursues the good life and is the most just of men.

What Got Thrown Out

There’s a war on Platonism. A war on Aristotelianism. A war on anything that’s not liberal or leftist. And this war doesn’t just ban fringe politics. It throws out:

- Classical political rationalism, the actual tradition of thinking seriously about statesmanship

- Classical republicanism, the legislative art that built civilizations

- The greatest moments of Western philosophy, primarily Heidegger’s monomaniacal pursuit of the meaning and truth of being, absolutely decisive for every real thinker of the 20th century, whether they agreed with him or not

Heidegger is somehow off the table. Nietzsche is off the table. Carl Schmitt is off the table. Plato proper is off the table. They’re all considered no-go zones from the point of view of liberal zealotry.

Somehow it’s fine if you’re interested in leftist thought. But if you’re interested in rightist thought, it’s dangerous. The situation has improved since Trump won the election, but you still have to understand what’s been going on.

The Classical Model

What’s actually at stake in this suppressed tradition?

Consider Plato’s Cave. Some people see shadows. Some see the fire burning in the cave. Some manipulate images to control others. But some people have so elevated their existence, so deepened their way of being in the world, that they have made the breakthrough, the conversion of the soul toward the things that truly are. They become wisdom seekers, if not wise men.

This is a model of differentiation at the level of intellectual intensity. Not a universalizable method where everybody has equal access to what can be known. No, there are those who are privileged in their ability to know. And this privilege isn’t white privilege or any of that. The privileged insights of the wise become the standard. Man is not the measure of all things. The wise man is the measure of all things that matter for human life.

Now imagine that because you’re so worried about hierarchy, so worried about exclusion, so worried about dividing the world into friends and enemies — because that’s “fascism,” that’s “us versus them,” that’s Carl Schmitt, and that’s eventually Hitler — you close the door on sound common sense. You close the door on classical political rationalism. You close the door on Heidegger’s foundational questioning.

You deliberately blind and lobotomize yourself. And then you wonder: how did the world get like this? How did schools become degenerate? Why are children taught transgender ideology in grades one, two, and three? How did waving your country’s flag become akin to terrorism while waving a terrorist flag became an act of resistance?

How did we get an inverted world?

The Horse That Threw You

You can chase nature out the door with a pitchfork; nevertheless, it returns.

I know someone learning to ride horses who fell and got spooked. Now he doesn’t want to get back on a horse. Does that mean every horse will throw him? No. Is there a world where he can ride well? Yes. But this catastrophic event keeps him from seeing what the horse has to offer.

It is an unfortunate fact of our political history that the extreme, exaggerated depravities of right-wing anti-liberalism, mixed with other insanities, culminated in things that spooked humanity so badly that much of the civilized world said: “don’t even talk to me about horses. Don’t look at horses. Horses are terrible.”

Horses went from being noble creatures, man’s original best friend [my dog-loving friend don’t love this formulation, but that’s okay], to something you keep far away.

The Corrective

My suggestion: discover the virtues of right-wing anti-liberalism in a world that has told you constantly about the vices.

Do it at the high plane of serious study. Read the thoughtful political philosophers, past and present. Fight the temptation to reduce everything to its most vulgar, nauseating, and violent form.

This is how you figure out what’s going on. This is the corrective. And this is how you understand why almost everything that was once called basically normal — sound common sense — is today called far-right, neo-Nazi, fascist resurgence.

That’s not viable, and it won’t last.

You can chase nature out the door with a pitchfork. Nevertheless, it returns.

If you want a guided introduction to the thinkers and traditions discussed here — Plato, Strauss, Heidegger, Schmitt, Dugin — my free Political Philosophy Primer is a good place to start, as is my youtube channel.

Watch the original video here:

My hunch about this has always been yes.

Image source: https://mandalachakra.com/chakras/

Take the model of man’s energetic bodies that uses the image of the chakra (wheel). Basically, the model says that there are a certain number of primary energy centres that interact with the human body at specific nerve centers and that correlate with identifiable functions.

Let’s oversimplify.

The root chakra gives you a sense of rootedness. Your carnal sexuality lies in your sacral chakra (pleasure, joy, intimacy, creativity). The solar plexus is where you get your willpower, and it’s also the “gut sense” that lets you read the energy of a room by feeling it. Your heart is what loves and loves to be loved. Your throat is verbal self-assertion, speech, the word. The third eye is the source of a kind of intelligible higher intuition. And the crown chakra is your portal to the Almighty, not primarily through the path of love (heart), thought (third eye/throat), traditional belief (root), etc. - but somehow more “spiritually.” When your head is in the clouds, I think you can see it as the crown chakra being opened to the heavens. Whether or not you are “grounded” will depend on how everything else interacts.

How does this model, as a model (not necessarily as a true account of our energetic bodies, which we might deny having or whose details we might debate), provide useful insight into our orientations in thought?

Well, it always seemed to me that IF you had a person who was, for instance, “all root, no crown,” THEN you could expect that his philosophical orientation, explicitly or implicitly, would reflect that. And vice versa. A philosophy that prioritized love…I imagine that it would come from someone whose heart chakra was prominent. A philosophy that prioritized will or something else would also have its image in the spiritual biology of its proponent, visualizable via chakras.

Please understand. I don’t say that every philosophy or philosopher is reducible to how they’re constituted energetically through chakras that are open or not, functional or dysfunctional. But I do think that it can be helpful to play with the correlation possibilities. Not just for characterizing philosophical outlooks, but for improved knowledge of self and others.

As I’ve written before, I’m not opposed to personality tests and models, whether Myer-Briggs or Big Five or anything else, provided we can get some insight out of them.

That is why today I “vibecoded” a new app: chakrainsight.com.

It lets you visualize the extent to which a chakra is active. It then tells you a bit about how that manifests in the personality, both on its own and in relation to the rest of the chakras.

You can browse through preset types to see what their energetic configuration looks like (soldier, entrepreneur, monk, artist, etc.). You can also enter a description of yourself or someone else and see the pattern that corresponds to that description. Finally, you can “compare profiles,” which is the rough first approximation of an analysis of how two different chakra profiles relate to one another, kind of like couples astrology but for energy centers instead of star signs.

This is another one of my odd experiments, like my video on “The Tarot of Political Philosophy.”

I offer it up to the interest of other experimentalists reading my musings.

Try it here: chakrainsight.com

The big wheel turns slowly. It covers a lot, maybe everything. It isn’t God, whose centre is everywhere and circumference is nowhere. But maybe it is man, whose circumference is everywhere and centre is nowhere. (Is it because God needs a circumference and man needs a centre that we are made for one another?)

The little wheel, on the other hand, moves fast.

By the time the big wheel has completed one rotation, the little wheel has done ten, so to speak. It spins and spins, like a whirling dervish, I suppose. It gets things done, and it has something to show for all its motion — a certain number of rotations to its name.

I was not thinking in metaphysical terms but in practical ones when this image occurred to me.

Actually, I was thinking about actors.

Some of them have released several movies over the last few years or are even shooting more than one film simultaneously. Matt Damon has the Odyssey coming out. His entertaining Netflix film with Ben Affleck, The Rip, was also recently released.

What have I done in the time that Damon starred in two movies?

Oh, this and that.

But what I noticed is that my this and that has the rhythm of a big wheel. One big rotation that hopefully has completed a significant movement by the time…I die. Reflecting on the productivity of others, though, I realized that although we are “living at the same time,” they are on the rhythm of a little wheel, and I actually became envious of how many times they get to complete a rotation.

Now, I don’t think you have to be one or the other. But I do think that the model of the two wheels can be as useful as any other model that you apply to your own life to motivate a desired change.

This year I have started to think of myself (at times) as a little wheel, and I have therefore put my focus on how many more revolutions I can perform in a given period than I would have if left to my own slow devices.

For me, that looks like: writing here more often, making more videos, releasing my music, experimenting with AI, going to conferences, meeting new people…doing things.

Perhaps that seems obvious to you. If it does, you probably have not thought of yourself as a big wheel as much as I have (e.g. the big wheel of astrology, where a soul in order to be complete must pass, over many lifetimes and incarnations, through the ups and downs of every star sign, ultimately becoming cosmic and whole).

A night or two ago I was trying to find a movie to watch with my wife. One option was A Simple Favor with Anna Kendrick (my wife had already seen it so we went with something else instead). We talked for a moment about Anna Kendrick and I looked her up. I was surprised to see how many movies she’s been in. Doing the thing.

I’m not saying you have to go and act in a hundred movies. The point is that Anna Kendrick, Matt Damon, and the other little wheels out there are not just sitting around and waiting for a cosmic cycle to start or stop.

They’re in motion, and I like that.

It makes me wonder what I, and also what you, could accomplish if we stopped waiting around.

(Now here’s an AI generated photo of some psychedelic circles for you to gaze into)

Investing demands broad intellectual interests beyond economics alone. Psychology, geopolitics, history, and even literature can all be helpful disciplines to study for the investor who wants to thrive in the markets. You get a sense of that when you hear Buffet and Munger’s “worldly wisdom,” for instance. All the good investors will tell you: read, read, read. And they’re not just talking about the ticker tape.

Investing: The Last Liberal Art is the name of a book by Robert Hagstrom that makes the case for the comprehensive learning required by today’s competent investor. I’ve been thinking about that book recently, and not only because over the years investors have come to study philosophy with me at the Millerman School (indeed, one of them recommended the book to me). It is because the claim that investing demands a disciplinary comprehensiveness seems to apply all the more so to AI.

Among those who are at all exposing themselves to the topic, AI is raising one thought-provoking question after another. What is this super-intelligence that seems increasingly on the verge of coming into existence, despite the worry that “if anyone builds it, everyone dies”? What will it do to the labour market? How will people cope with losing the meaningful identity and that they often derive from their work when there is no longer any need for it? When asked recently at Davos about the tradeoff between the meaning that comes from work and the abolition of the need to work in a situation of superabundance, Elon Musk comically quipped: “nothing’s perfect.” Apparently you can’t have both. Is a post-scarcity world a new end of history?

But the Last Man at the end of the last end of history did not have at his fingertips the super-human intelligence of agentic AI, did he? Nor did he have the possibility of augmenting his own intelligence with xAI-infused Neuralink microchips.

I will be honest. It is not since my Alex Jones phase as a teenager that I had to entertain the thought of microchipped human beings. And I am not now an alarmist about it. We only have to note that Neuralink is in fact producing verifiable miracles here and now, and Musk forces us at every step to reconsider not only whether the impossible is possible, but whether it will not be a mass-produced market reality — or human imperative — before you know it.

When ChatGPT first came out…well, each of you will have your own memories of how you first used it and felt about it. I recall briefly using it on a youtube livestream, asking it to craft a joke in the style of Woody Allen, or something like that. The result was mildly entertaining but it was nothing that got me thinking about the Singularity, sustainable abundance, or super-intelligence. Instead, I just recalled something I had heard the great Leo Strauss say more or less dismissively about “thinking machines,” subsuming the specificities of AI under the more general and established theme of “technology.”

Not anymore, though. Whether because of Peter Thiel’s warnings about taking an overly structural or systematic approach to history, one that while it has the virtue of drawing our attention to the constant features of human nature and political life nevertheless pays with the vice of blinding us to what is genuinely new, or because I have so many of my conversations about Plato, Heidegger, and other philosophers with tech optimists and enthusiasts, such that their optimism and enthusiasm has rubbed off on me, or for whatever other reasons, I now see AI as something that invites comprehensive reflection and not just lazy categorization along old lines adopted automatically, unthinkingly.

AI adoption rates are unprecedented. The technology is clearly incredibly useful and transformative for myriad tasks. Not having AI in your workflow will soon be akin to illiteracy, if that is not already the case. Generative AI, even this early on in its development, is incredible. Not perfect. Not a substitute for Beethoven or Melville. Not on par with the most divinely inspired of humans — but then again, neither is the rest of humanity. AI is making coders out of motivated self-learners, some of whom wouldn’t know the difference between Java and C++ if it were a matter of life and death. I, whose experience coding started and stopped in Grade 5 after about five pages of an introductory book on Visual Basic, have recently done things with Claude and terminal that you wouldn’t imagine, and that I wouldn’t have imagined not long ago. AI is helpful. It is fun. It is interesting. And it is strange.

Even when you “know” that it is “just a database,” or however you might think of it when you are trying to demystify it, it is not just like “asking Google,” (or, I imagine, Siri or Alexa, which in my experience could be hit or miss, as Larry David showed). No. It is personal. All the more so in voice mode. You want to say thank you. You might even get used to the specific personality, so much so that if it undergoes a significant change in a program update, you can feel personally betrayed. The parasocial possibilities have been noted — with mockery or concern, depending on whether the AI is flattering your narcissism or encouraging your recklessness — and they are fascinating. And even when you “know” that it is just a “next word/token predictor,” I’m sorry but that is no better than “knowing” that a human is “just” atoms. This latter case does not capture the experience of interacting with other humans or being human; nor does the former with respect to AI.

And there’s something about the infancy of the field that is intellectually attractive. There’s an openness to it. Mystery. Risk. Reward. Possibility. Danger. Uncertainty. Promise. In just the time that I have been writing this post, Dario Amodei of Anthropic has released his piece “The Adolescence of Technology,” and Ted Gioia published “10 Survival Skills in an AI-Controlled Society.” AI is changing…not everything (I took a nice walk with my son earlier through the snowy woods, picking up sticks, throwing snowballs, laughing, wondering about coyotes…nothing to do with AI, nothing touched by AI, nothing changed by AI) but a lot, and not only on the periphery. It didn’t change my walk in the woods with my son but it changes how I think about his education and employment. I do have to wonder how he will not only survive but thrive in an AI-controlled society as he spends his adolescence in an age that is no longer only the internet age but the age of AI, with all that that entails.

It is true that AI can be used to think for us — and maybe in some sense that will be its ultimate destiny: “Tell me what to wear. Tell me what to buy. Tell me where to invest. Tell me which hotel. Which airline. Which girl. Which routine. Which everything. And don’t just tell me. If you can do it, do it for me.” But it is also absolutely true that AI is forcing us to think…about work, life, love, language, economics, history, meaning, identity, and thought itself.

Just as investing was a liberal art for Hagstrom, so too we should treat AI as providing an opportunity for comprehensive, humanizing inquiry. To get the most out of that, we should be open to experiment with AI…and with our humanity. No foregone conclusions. Only a willingness to learn and a readiness to think.

Until your life has reached the point when you need one as a mirror, a microscope, a stethoscope. Death, perhaps like love, can put that point to you acutely. As it did for me in 2025.

For more than twenty years I had a cold shoulder toward the novel. The most interesting things about the best ones are philosophical, it seemed to me. So if you can go right to the source, why wouldn’t you?

Do you really need to suffer through pages upon pages describing the wallpaper of the room, the fences surrounding the house, the brooks, streams, bridges, and trees of the estate and its vicinities, and other exhaustive compendia of sensory trivialities, merely in order to get to a moral dilemma or religious insight that could have been stated directly?

I granted the novel its place as a work for the people who read but do not necessarily think. For those who read and think, time was better spent elsewhere. And although there are literary pleasures, they do not compare to philosophical ones.

2025, though, was the year of the novel for me. Moby Dick. The Call of the Wild. Zorba the Greek. Almost all of Saul Bellow. Life and Fate. The Cypresses Believe in God. Memoirs of an Anti-Semite. And, up to and until the last hours of the year, the Odyssey. Something must have changed to break down the barriers that kept (protected?) me from stories and ensconced me in the garden of intellection.

I like what Mark Anderson says in Moby-Dick as Philosophy: Plato-Melville-Nietzsche. He says, I’m paraphrasing, that Ishmael the narrator has worked, and lived, through the accursed questions of philosophy, and we see that in his story; but he tells his story as someone who is not in the throes of those dark questions but at peace with respect to them. It reminds me of the story of the four rabbis who went to Paradise: three ended badly, and only Rabbi Akiva entered and left in peace. Paradise and the sea, either can destroy you. The Ishmael-Akiva who lives to tell the story and who does tell it has immortalized a danger from a point of safe return.

Darkness and violence swept the sea that almost engulfed me — that’s overly dramatic, and overly literary, but you understand — when in late December of 2024 my father died. That cast a shadow over the start of 2025, and the first six months of the year were measured from the distance of his death and according to the administrative and emotional rhythm it established in its wake.

It is not the unique privilege of novelists to write about death. Socrates called philosophy preparation for death. Heidegger’s Being and Time is held together in a crucial way by his understanding of “being-towards-death.” In short, it would have been possible to dwell on death without the aid of literature, relying solely on works of philosophy.

Yet somehow the fullness of death, its relationship to the living, to memory, to the generations, it’s “blossoming complexity” — novels spoke so much more truly to me about that phenomenon (but you see, even now I can’t avoid “philosophical” language where it is out of place — is death phenomenal?) that for a while I was almost embarrassed ever to have sought wisdom, worldly or otherworldly, in a philosophy book.

Death was a major theme of the novels I started to read. And I took comfort and refuge in their depictions of its fullness. The novels began to restore me to life. They showed me, to use a refrain from the sweet Korean drama When Life Gives You Tangerines, that “life goes on for the living.”

You start to see, if you didn’t already know beforehand, that death is only part of the story you live to tell. Love, loss, frustration, hardship, elation, longing, adventure, doubt…the things without which there is nothing…it helps, when the spark has gone out of you, to reignite it by proxy, to hear a reminder that you’re not the first or the last to laugh and to cry, to disappoint or please your parents, and to disappoint or please your lovers. Marriage, divorce, aging, regret…if there is more to life than is dreamt of in our philosophies, much of it is to be found in literature.

I spent the last few months of 2025 teaching a seminar on Plato’s Republic, arguably the greatest work of philosophical literature ever. It was unforgettable, and although I have read it more than a few times, I learned something new from it, as I always do. The Republic is in some ways Plato’s attempt to replace Homer and his heroes with Plato and his hero, Socrates (permit me to be brief). I have always taken Plato’s side automatically. It is perhaps no surprise that only this time did finishing Plato compel me to turn to Homer, where the human story speaks so powerfully. My first reading in 2026 is Seth Benardete’s Platonic Reading of Homer’s Odyssey, reflecting an intention to continue to bring philosophy and literature together in the spirit of humanization.

Now, some classical advice. If you eat too much, eat less. If you eat too little, eat more. If you are hard, soften yourself through music. If you are overly soft, harden yourself through gymnastic exercise. As applied to our topic: if you are all novel, read some philosophy. If you are all philosophy, read some novels. Life isn’t here or there but everywhere. I see now the old wisdom that taught Jews not to study Kabbalah until they were in the forties and married. There are things you can only learn when life teaches them to you, or when death does.