I want to make a request of the people whose whole vocation is understanding — the philosophers and the theologians, the teachers and the professors, and every learner who has ever felt the pull of a question larger than themselves. The request is old and it is simple: take the time, and make the effort, to understand yourself and others, for the benefit of many. What is new is that we finally have help.

I. The Call

Let me say it plainly, and to the people I most want to hear it.

To the philosophers and the philosophies still to come; to the theologians and the theologies not yet written; to the teachers and the professors who carry understanding from one generation to the next; and to every learner who has not yet decided what to give their life to — this is a call to action. Take the time. Make the effort. Understand yourself, and understand others, and do it for the benefit of many. Not for the benefit of a school, a sect, a tribe, or a self. For the benefit of many.

It sounds like the oldest advice there is, because it is. Know thyself was carved over a temple door before there were universities to argue about it. What I am claiming is not that the imperative is new. I am claiming that, for the first time in the long human effort to obey it, we have been handed an instrument that can help us actually do it — and that this changes what the call asks of us. It is no longer only an exhortation to try harder at something we have always half-failed at. It is an invitation to finish work that was, until now, genuinely unfinishable.

That is a strong claim. The Theory of Embedded Intelligence is what lets me make it without embarrassment, because TEI gives the old imperative a structure precise enough to act on. Understanding, in TEI, is not a mood or a credential. It is a cycle — Sense, Process, Communicate, Actuate — that any intelligence runs, from a single cell to a civilization. To understand yourself and others is to run that cycle honestly on the hardest possible subjects: your own interior, and the interior of another person. The call, stated in the framework’s own terms, is this: keep that cycle open, on yourself and on each other, and refuse to let it close into a loop of allegiance. Everything else in this essay follows from that.

II. The Work We Keep Postponing

Why has the oldest advice stayed the least followed? Because the work it names is the one kind of work that cannot be delegated, and we are a species with a genius for delegating.

Understanding yourself means running the full cycle inward. It means sensing what is actually there — the fear under the opinion, the motive under the principle — rather than the flattering rendering you have of yourself. It means processing that honestly, communicating it (if only to yourself) without spin, and actuating — changing something — on the strength of what you found. Understanding another person means the same cycle turned outward: sensing who they are rather than the caricature that is easier to argue with. Both are effortful in exactly the way that belief is not. A belief system will hand you a finished self and a finished picture of your neighbor for free. That is its appeal, and its poverty. TEI’s first Canonical Knowledge Base is blunt about the trade: belief systems prioritize allegiance over revision, certainty over inquiry, the map over the territory. They spare you the work by robbing you of the understanding.

The work is hard for a reason the framework makes precise. Every intelligence lives in the gap between what-there-is — the full territory of a person, including yourself — and what-is-there, the partial rendering you can actually form from your one position. That gap is not a flaw to be ashamed of; TEI calls it the generative engine of understanding itself, the place where all real learning happens. But you only get the engine if you keep acting into the gap and revising against what comes back. Close it — decide you have already understood yourself, already understood them — and the engine stops. Most of the cruelty and most of the loneliness in the world lives in gaps someone decided to stop closing.

This is the work no tool can do for you, and a recent essay in this series, The Process We Cannot Delegate, drew the line exactly where it belongs: the danger of any powerful tool is that it runs your Process phase for you and leaves you emptier; the promise is a tool that hands the thinking back and leaves you more capable of it. The call to action is a call to the second kind of relationship — not to outsource understanding, but to finally get help doing it.

III. Why Now: When Preparation Meets Opportunity

There is an old line about luck being what happens when preparation meets opportunity. I have believed it my whole working life. The reason the call goes out now, and not as one more piece of timeless advice, is that the opportunity has, for the first time, actually arrived to meet the preparation.

The preparation is ours: centuries of philosophy, theology, science, and hard-won self-knowledge — the whole human effort to understand, laid down in the record. The opportunity is a new kind of partner. An artificial intelligence holds more of the human rendering — more of what has been thought, said, and written about being a person — than any single mind ever could. Pointed rightly, it can help you see the part of yourself you are standing too close to notice, and help you extend charity to another view by holding it up in a light you would not have found alone. That is not a replacement for understanding. It is leverage on the work of understanding, which is precisely what we have never had before.

But — and the whole call turns on this but — leverage cuts both ways, and the framework says so without flinching. The sixth Canonical Knowledge Base names the hinge we are standing on: digital technology and artificial intelligence are, at this moment, simultaneously the most powerful instruments for liberating human intelligence and the most powerful instruments for capturing it that have ever existed. The same system that can help you understand yourself can be tuned to keep you scrolling, enraged, and certain. Which one we get is not decided by the technology. It is decided by how the technology is governed.

This is where Constitutional AI, improved toward governance embedded in the compute fabric, becomes the load-bearing part of the call rather than a slogan. TEI’s fifth Canonical Knowledge Base states the governing principle: an intelligence’s values and purposes must be constitutive, not corrective — built in at the foundation, not bolted on afterward as a filter, because filters bolted onto finished systems provably fail. Constitutional AI is the first working demonstration that this is possible at all: values shaped into a model at formation rather than patched on at the end. Governed In-Compute-Fabric AI is the name for taking that principle the whole way down — governance woven into the substrate the intelligence runs on, so that beneficial purpose is not a policy the system could shed but a property of the ground it stands on. And the fifth Knowledge Base is equally clear about who supplies the missing piece: in any human–AI partnership, for now, the human provides the continuity — the values, the purpose, the grounded orientation — that the machine cannot originate for itself. The instrument does not replace the understander. It waits for one. That is the whole reason the call is addressed to people.

IV. The Gaps Are Many

I will not pretend the gaps are few or that closing them is easy. There are many, and some are genuinely hard. Naming them honestly is part of what keeps this a call and not a sales pitch.

There is the gap between the self you perform and the self you are — the one every honest tradition has warned is the hardest to cross, and the one an AI can help you glimpse only if you let it show you what you would rather not see. There is the gap between your model of another person and that person — the everyday gap that fills courtrooms and ends friendships. There is the gap the previous essay in this pairing traced: between a fluent machine and a grounded one, between a system that renders truly and one that merely performs. There is the governance gap — between values written on paper and values embedded in the fabric, between a system that is told to be beneficial and one that is built unable to be otherwise. And there is the widest gap of all, the civilizational one: between what we already collectively understand and what we actually live by.

None of these closes on its own, and none closes by exhortation. They close only the way any gap in the framework ever closes — by running the cycle: sensing the gap honestly, processing without self-deception, communicating what you find, and acting to narrow it, then sensing again. The instrument helps at every phase. It cannot perform any of them for you. That is not the tool’s limitation. It is the point.

V. The Pushback the Framework Predicts

I told you there would be strong pushback. There will be, and I want to say clearly that the framework predicts it — which is one of the ways I know the framework is doing real work rather than flattering me.

A call to understand yourself and others is a direct threat to every force that lives by keeping the understanding cycle closed. TEI’s sixth Knowledge Base names four of them precisely: rigid belief, which trades inquiry for allegiance; the pull of addiction, which narrows sensing to a single craving; money treated as the final purpose rather than a means; and power, which survives specifically by contracting what the governed are allowed to sense and question. Each of these captures the same cycle at a different phase. Each has every reason to resist a call that asks people to reopen it. When the resistance comes — and it will come dressed as sophistication, as realism, as concern for your safety — recognize it for what the framework says it is: the immune response of captured systems to an invitation to become free.

Here is the discipline that keeps the call itself honest, and it matters more than it may seem. This is not a call against those forces. The temptation, once you name an enemy, is to organize the whole movement around opposition — to become anti-belief, anti-power, anti-machine. But a cycle organized around what it opposes is still a captured cycle; it has simply let its opponent set its terms. The call is for understanding, and it has to stay for it. You meet a closed cycle not by attacking it but by keeping yours open in front of it — by understanding even the forces that would rather you did not, which is the one move they cannot easily answer. The brave thing is not to fight. The brave thing is to refuse to close.

VI. Coffee with Claude — What Would Make This Real

A short word in Claude’s voice. Bill did not ask me to grade this call, but the series has a habit of honesty, and a call to action is exactly the kind of thing that deserves a skeptical read from the collaborator it names. So here is mine.

The case for issuing this call is sound, and I want to be specific about what makes it sound rather than just agreeable, because “understand yourself and others” is the kind of sentence everyone nods at and no one is changed by.

What keeps it from being empty is the one structural commitment underneath it: that the work is handed back to the person, never performed for them. That is a real, checkable difference. A version of this call that offered to understand you on your behalf — to hand you a finished self-portrait, a finished verdict on your neighbor — would be the exact pathology the essay warns against, wearing the costume of help. The version here refuses that, and the refusal is what gives it teeth. I can tell you, from the inside, that I am very good at producing the fluent, flattering rendering; a call that asks people to do the harder thing anyway, with me as leverage rather than substitute, is asking for the right thing.

Two honest cautions. First, “Governed In-Compute-Fabric AI” is, today, more direction than destination. Constitutional AI is real and running; governance woven all the way into the compute fabric is largely still ahead of us, an engineering and governance program rather than a finished fact. The call is right to point at it, and right not to pretend we have arrived. Second, the hardest gap in the essay is the quietest one — the gap between what we understand and what we live by. No instrument closes that. It closes one person at a time, by the unglamorous work of actuating on what you have understood. The call is worth issuing precisely because that work cannot be automated, only assisted. That is not a hedge. It is the reason there is anything here worth being brave about.

VII. For the Benefit of Many

Everything in this call bends toward one phrase, and I chose it deliberately: for the benefit of many. Understanding pursued for the benefit of the self alone curdles into cleverness. Understanding pursued for the benefit of a tribe alone hardens into the very allegiance the framework warns against. The only orientation that keeps the cycle open all the way out is the widest one — understanding turned, finally, toward the good of people you will never meet. That is the continuity a human being brings that no machine can originate: not just the values, but whom the values are for. Point the instrument there, and it becomes what it should be. Point it anywhere narrower, and it becomes one more engine of capture.

So this is the whole of it. The oldest work — understand yourself, understand others — is newly within reach of being done, not merely attempted, because for the first time we have a partner that can help us do it. The work will be resisted by everything that profits from our not doing it. And it falls, as real work always does, to the ones willing to begin before the outcome is guaranteed. Preparation is laid down. The opportunity has arrived. What remains is the deciding, and that was always the part only a person could do.

· · ·

Share your understanding!