The Vatican Named the Moral Stakes.
Here Is the Architecture.

I want to tell you about a single sentence in a new papal encyclical on artificial intelligence — a sentence I have, as it happens, been living on the inside of my whole life.

I am an autistic leader. I say that out loud, as a senior leader, in organizations and in public. And I am leading with it because it is the entire vantage I bring to this, and the vantage is a narrow one. For the theology, the Church history, the full ethics of these systems, there are people who do that work far better than I could, and I will point you to them. What I can give you is the one thing they can't. This encyclical makes a claim about dignity and cognition — about whose way of thinking counts, and what happens to people whose minds work differently when a powerful new technology arrives. Everyone else is describing that claim from the outside. I am reporting from inside the cognition itself. It is the only vantage I have, and it is worth your time precisely because it is narrow.

Now, saying "I am an autistic leader" out loud does not always land well, and I want to be honest about that. I have been told the label is just an excuse for personality flaws. I have been told it explains away a deficit of empathy. And I have been told, more gently, that it is a thing to be managed — that the disclosure is a liability I am choosing to carry.

I keep saying it anyway. And the reason has almost nothing to do with me.

Because by most measures, I am the easy case here. I am verbal. I am employed. I am in the room where decisions get made. If you were ranking the people whose dignity is actually on the line in the conversation I am about to have, I would be near the bottom of that list — I am the luxury layer. The neurodivergent people whose lives are most shaped by how organizations decide to treat different kinds of minds are affected in ways I will never fully feel. I get to talk about this from a position of safety that most of them do not have.

And I have come to believe a position like that has exactly one good use. You spend it. You spend it on what it can actually reach. I chose not to hide the thing about myself that I could have hidden, and the only thing that makes that choice worth anything is whether I use the safety it gives me to push on something real.

The thing I am pushing on, in my lifetime, is dignity. And I mean something very specific by that word. I mean the plain claim that a person's worth is not a reward their way of thinking has to earn.

Hold onto that sentence, because it is the spine of everything that follows. Dignity isn't earned. It's the precondition.

Now here is what happened, and why I am talking to you about it this week.

A Pope just named that exact claim. In theological language, with the full institutional weight of the Catholic Church behind it. And he did it in the context of artificial intelligence.

But here is the part that made me sit up. He is not the first to name it, and he is not even the second. A United Nations convention put the same claim into international law twenty years ago. And a research field is, right now, measuring the same claim as economic infrastructure — something called brain capital. Three completely different vocabularies. Theology, law, economics. None of them coordinated with each other. None of them is citing the others.

And they are all pointing at the same architecture.

I want to be careful about my place in this. Putting myself next to a Pope and a UN treaty would cheapen the whole thing, so I won't. I am the smallest voice in that convergence — and, I think, the only one reporting from inside the cognition the other three describe from the outside. That is the contribution: the location of the voice, not the size of it.

Let me give you the detail that convinced me this was worth an essay and an episode, because it is too specific to be an accident.

On May 15th, 1891, Pope Leo the Thirteenth signed an encyclical called Rerum Novarum. That document addressed what they called the social question — the dignity of workers, the rights of labor, the moral stakes of an economy in upheaval — in the context of the first great industrial revolution. The steam engine. The factory. The machine arriving into human work.

On May 15th, 2026 — 135 years to the day — Pope Leo the Fourteenth signed an encyclical called Magnifica Humanitas. Same date. And he addressed the same question, the dignity of human beings and their work, in the context of the AI revolution. He took the name Leo on purpose, and he explained why himself. He said he took it "mainly because Pope Leo the Thirteenth, in his historic encyclical Rerum Novarum, addressed the social question in the context of the first great industrial revolution." And the Vatican press office added that the choice was, quote, "in reference to men and women and their work, also in the time of artificial intelligence."

He presented it on May 25th, alongside a co-founder of the AI company Anthropic.

Same name. Same date. Same question. The architecture being defended did not change in 135 years. The only thing that changed is the technology forcing the defense — twice. The steam engine, then the algorithm.

I want to be clear about what kind of reading this is, before I go any further. It is a structural reading. The only part I can speak to from the inside is what this claim does once it stops being a moral principle in a document and starts colliding with the ordinary machinery of how organizations route cognitive work. That collision is where I live. The rest of it, I am reading the way anyone can read it.

So let me build it for you, piece by piece. There are really four moves here, and I am going to walk you through all four.

The first move is the word everyone is getting wrong.

The verb at the center of this encyclical is "disarm." And the moment people hear a Pope say "artificial intelligence must be disarmed," the dominant read kicks in — disarm means reject. Weaken it. Retire it. Stop using it. The Church is anti-technology, the Pope is against AI, that whole reflex.

That read is wrong, and it misses the entire architecture. So let me let him define his own term, because he does, very precisely. Here is the call to action, in his words:

"Artificial Intelligence now demands to be disarmed, freed from logics that turn it into an instrument of domination, exclusion and death."

And then, in the same chapter, in almost the same breath, he tells you exactly what he does and does not mean. He says: "To disarm AI means freeing it from the mentality of 'armed' competition." And then, even more directly: "To disarm does not mean rejecting technology, but preventing it from dominating humanity."

Sit with that for a second, because it changes everything downstream.

Disarm is not a refusal move. It is a design move. The work is not to take the capability away. The work is to design the architecture around the capability so that the capability cannot wound dignity — at any level it touches. That is a completely different instruction than "stop using AI." It is an instruction to build something.

And here is the honest part. The Vatican did not invent that work, and neither did I. The encyclical restated it — with enough cultural authority to make it impossible to keep ignoring. What I can add is narrower and more useful than the moral framing. I can tell you what that design move actually looks like, concretely, at the four levels where organizations route cognition. That is the rest of this episode. But first I have to show you that the claim itself is older than the encyclical.

That is the second move. Because the Pope did not invent this either.

Twenty years ago, the international community wrote it down. The United Nations Convention on the Rights of Persons with Disabilities was adopted by the General Assembly on December 13th, 2006, and it entered into force in May of 2008. And I want you to hear the scale of this, because it matters. 192 state parties, plus the European Union, have ratified it. It is one of the most-ratified human rights treaties in the entire history of the UN. This is not a fringe document. This is close to the whole world signing the same page.

And the word that anchors that document is the same word the Pope reached for two decades later. Dignity.

Article One says the purpose is to "promote respect for their inherent dignity." Article Three, the very first general principle, leads with "respect for inherent dignity." It is load-bearing. It is the foundation, not a footnote.

But here is the structural move inside that convention, and this is the part most people miss. The convention's move was not "accommodate disabled people." That sounds like the same thing, but it is the opposite thing. "Accommodate" assumes a system that was designed for one kind of person, with adjustments bolted on afterward for everyone else. The convention inverted that. It said: design the system so that dignity is the precondition — the thing the architecture begins from — not the outcome you reach if everything goes well.

Think about the difference. The older model treated dignity as something an organization earned by not failing. Don't discriminate, don't exclude, and you get to call yourself dignified. The convention said no — dignity is the starting architecture. It is what you build from, not what you arrive at.

And that is exactly the Pope's framing in 2026, restated for a new technology. Dignity comes before productivity. Not the other way around. Same architectural posture, twenty years apart, in two completely different vocabularies.

Which brings me to the third move, and this is where my own world comes in directly.

The convention is making a claim about disability. The neurodiversity movement is making a claim about cognition. And I want to say plainly: those are the same claim, expressed twice.

Here is the neurodiversity claim, in the language this platform uses. Different nervous systems produce different cognitive architectures — shaping how people think, sense, process, and decide, across mind and body. Not broken brains. Variation to be designed for, not filtered out. The convention said it in the vocabulary of international law and human rights. The neurodiversity movement said it in the vocabulary of lived cognitive experience. They are not two movements pointing at two adjacent problems. They are one principle, spoken in two languages.

And then there is the third language. Economics. Because what happens when you take that principle and try to quantify it? You get what researchers and the World Economic Forum now call the brain economy, and its central unit, brain capital. Brain capital is brain health plus brain skills — the framing of human cognition as measurable economic infrastructure. Capital you can invest in, capital you can deplete, capital with a return.

And that quantification is a double-edged thing, and I want to be honest about the edge that cuts the wrong way. Brain capital is dignity-preserving infrastructure only when it is designed to be. Measure it the wrong way, and it collapses straight back into Taylorism with a neuroscience vocabulary. A more sophisticated stopwatch. A more elegant way to rank people by their output, while keeping the actual architecture — the thing that produces the variation you are measuring — completely invisible. The vocabulary of brains, in the service of the old industrial sorting.

So you have three vocabularies, all pointing at one principle: dignity isn't earned, it's the precondition.

And then the third movement arrives. Because if the neurodiversity movement made the claim visible, and the brain economy made it economic, the Age of AI is what makes it urgent. AI is the pressure that makes the architectural choice impossible to keep avoiding. Here is why. Every single AI deployment is a routing decision. It routes cognitive work between human judgment and machine execution. That is what it is, underneath all the product language. And you can do that routing two ways.

Route it without an architecture, and AI absorbs dignity right along with the work. It flattens the variation in how people think down into prompt-engineering style. It treats human judgment as friction — as the slow part the technology is here to remove.

Route it with an architecture, and AI absorbs the work that does not require dignity, and it protects the capacity for the work that does.

That is the disarming work the Pope named. It is not a slogan. It is the deliberate design that decides which of those two routings actually happens in your organization. It is the work the brain economy was always going to require eventually. AI is just what made it impossible to keep deferring.

Now I get to the fourth move, and this is the part I most want to be careful about, so let me set it up clearly.

I am about to take four specific claims from the encyclical and pair each one with a structural consequence inside an organization. And I need to name what this is and what it is not, because the line matters.

These are correspondences. They are not endorsements. The Pope is naming a moral principle. He is not endorsing me, he is not endorsing my work, he has never heard of me, and nothing I am about to say should be heard as "the Vatican backs these frameworks." What I am doing is reading what the principle forces — structurally — once it stops being theology and meets the way work actually gets routed on an organization's floor. The forcing is what is interesting. Not the blessing. There is no blessing. There is just an architecture that the premise makes unavoidable.

Four claims. I will walk you through each one. Listen for the same move each time: he names the moral principle, and the architecture does the rest on its own.

Claim one. The Pope writes, "We cannot consider AI to be morally neutral."

Here is what that forces, structurally. Every AI deployment sits on top of an existing cognitive architecture. AI does not arrive into an empty room. It arrives into an operating model that already has assumptions baked into it. Assumptions about whose thinking counts, which kinds of cognition are visible, which moves get rewarded, and which signal gets quietly treated as noise. So when you drop an AI system on top of that, you are not adding a neutral tool. You are amplifying whatever architecture was already there. And let me make that concrete with a number that has been on my mind. Research out of Harvard, Wharton, and Deloitte earlier this year found something like a 93-to-7 split — 93 percent of the investment and attention flowing to the technology, about 7 percent to the human cognitive side of the equation. That lopsided split is exactly what happens when the substrate goes undesigned. The money flows to the technology because the technology is legible. You can see it, buy it, point at it. The human cognitive architecture stays invisible because nobody ever mapped it. So the Pope's claim that no AI deployment is morally neutral lands first, before anything else, as an organizational claim. And the diagnostic is brutally simple. Was the cognitive architecture underneath your AI ever deliberately designed? Or was it inherited from the old industrial operating model and never once re-examined?

Claim two. The Pope offers a binary image. He says the choice is, quote, "either to construct a new Tower of Babel, or to build the city in which God and humanity dwell together." And he is specific about what went wrong at Babel — he says it "eliminated diversity" by forcing a single language onto a world that did not have one.

Here is what that forces. The translation tax is the cost of pretending that one cognitive style is universal. The single language at Babel was words. The single language in most organizations today is not words anymore — it is a single cognitive style, treated as the default, the professional one, the right one. And anyone who processes the world differently pays a tax. They have to reword their actual thinking into the dominant style before it is allowed to land. And here is where AI compounds it, badly. AI deployed with no communication architecture optimizes for the style its training data over-represents. Fluency, social smoothness, the kind of predictive sociability that reads as polished. And it treats divergent cognition as noise to be cleaned up. Now let me make this one something you can actually hear, this week, in your own organization. Notice who, in your next meeting, visibly reworks their thinking into the dominant register before they speak. The pause. The translation. And notice how much of the actual signal thins out in that rewording — how the sharpest part of what they were thinking gets sanded down to fit the room. That delay you are watching is not a personality trait. It is a design choice that nobody made on purpose. There is an architecture that preserves the signal instead of refining it away. The platform calls it the Cognitive Translation Protocol. But hear me clearly: the tax is imposed by the missing architecture. Not by the absence of any particular framework. The framework is just one answer. The tax is real whether or not you have ever heard of the answer.

Claim three. This is the one closest to the spine. The Pope writes that every human being has dignity, quote, "simply by virtue of existing, of having been willed, created and loved by God."

The theology is his. The structural consequence is organizational, and it is exact. Dignity is the routing principle. Productivity is not. Most AI strategy routes cognitive work by one question: what can we automate? Output is the sorting key. The Pope's claim inverts the sort. Dignity comes before output; it is not earned by output. And — this is the part I need you to hear — that inversion is not a nice value you bolt onto your strategy afterward. It is what the architecture forces the moment you accept the premise. Because cognitive integrity work, deliberate judgment, principled dissent, the refusal to act under uncertainty, has to be preserved. Not because the worker earned the right to it. Because the substance of that work degrades the instant you route it for output. Route judgment for speed and you do not get faster judgment. You get something that is no longer judgment. And here is how you derive where your own organization sits, without hiring anyone to audit it. Look at your last AI deployment. Ask one question: which work was it allowed to absorb? If the sorting key was "what can we automate," rather than "what must stay human because the substance requires it" — then your architecture is optimizing output and calling it strategy. The routing logic that takes the inversion seriously is the AI Cognitive Strategy Matrix. But again, the inversion is forced by the premise before any framework gives it a name. Dignity isn't earned. It's the precondition. Even for the work.

Claim four. The Pope writes, "A more moral AI is not enough if that morality is determined by a few."

What that forces is a claim about decision rights. They have to be distributed. There is no single oracle. And the structural reason a single oracle fails is not that the person at the top is malicious. It is that a single point of moral arbitration has no error-correction. Encoded ethics — however careful, however well-intentioned — becomes domination the moment the encoding sits in one hand, because nothing downstream of it can dissent. There is no mechanism to say "you got this one wrong." And the Pope's own counter-image is precise here. It is Nehemiah: the wall of Jerusalem rebuilt by, in the text's phrase, "the shared responsibility of all." Each section owned by someone. No central oracle handing down the whole design. The organizational version of that is exact. You separate three things that collapse together in most companies: signal, decision, and accountability. You let AI sit inside each layer where it earns its place, synthesizing the signal, surfacing the options, modeling the consequences, while the decision itself stays human, the accountability stays a named person, and the authority stays distributed across the structure. And here is the test you can run on your own last decision. Your last AI-assisted decision. Could you name who synthesized the signal and who owned the call, as two different, named people? If those collapse into one node — whether that node is a human or a model — then you have concentrated exactly the authority the encyclical is warning against. The architecture that refuses that collapse is DecisionOS. But the refusal is forced by the structure. Not invented by the framework.

So that is the four. And I want to pull them back together, because the point is not that there are four separate problems with four separate tools. The point is the opposite. These four dimensions — the organizational substrate, the communication tax, the routing principle, the distributed decision rights — are not four frameworks aimed at one problem from four angles. They are four faces of a single architecture. The encyclical's claim lands in all four of them because the moral stakes are not sitting at one tidy level of organizational life. They are present at every single level where the cognitive work actually happens.

Let me bring this in for a landing, because I think the honest conclusion here is also the humble one.

None of this changes the work. It corroborates it.

Think about the through-line again. 1891 — the steam engine. The dignity of workers against the machine. 2006 — the accommodation model gets retired and replaced with a dignity-first one, in international law. 2026 — the algorithm. Three different technologies. Three different vocabularies — theology, law, economics. And the same claim every single time. Dignity is the precondition. Not the reward.

The encyclical is the loudest voice to ever say it. It is not the first. And it will not be the last. It does not need me to agree with it, and it certainly does not need me to co-sign it. What it does, by being so loud, is make the claim unignorable for a moment. And I am trying to use that moment for the one thing my vantage is good for.

Here is the sentence I would leave with you. A person's worth is not a reward their way of thinking has to earn. Dignity is the precondition, not the output.

And here is the part I want to be careful and honest about. I have spent years specifying what this design actually looks like at each of those four levels — the lived vantage is what made the failure modes impossible to un-see, because I was the one paying the tax. Three of those specifications have names. The AI Cognitive Strategy Matrix, for the routing. DecisionOS, for the distributed decision rights. The Cognitive Translation Protocol, for the signal preservation. They sit inside a broader organizational architecture — the thing I address as a whole in a keynote called The Undesigned Mind. But I want to be precise about what those are. They are one operational reading. They are not the answer the encyclical demands. The encyclical demands dignity. These are just tools for meeting that demand. The demand is the thing. The tools are negotiable.

So here is the exit, and then one last thing.

If any of this resonated, the full essay is up at theautisticleader.ai — it is called "The Vatican Named the Moral Stakes. Here Is the Architecture." And if you want this kind of thinking week to week, the newsletter is at theautisticleader.substack.com. It is free, there is no pitch cadence, and you can unsubscribe any time. I would be glad to have you.

But I do not want to end on a subscribe button, so let me end where it actually matters.

You do not need me, and you do not need a Pope, to tell you whether your organization is doing this work. You can check it yourself, today. Look at your last AI deployment, and ask two plain questions. What was it allowed to absorb? And whose judgment did it quietly route around?

The architecture was already there before you asked. It was either deliberately designed, or it was inherited from the old industrial model and never once examined. That is the only real fork. The encyclical did not create the choice. It just made the choice impossible to keep ignoring.

The architecture is yours to design.

Thanks for listening. I'll see you next week.