The Vertigo of the Honest Engineers
Lagune near Burano, Venezia. May 2026
What a theologian observes from inside the AI industry – and the question he doesn't ask
A week ago, I wrote on this blog about the alliance between the Vatican and Silicon Valley that has formed around the encyclical Magnifica Humanitas. Christopher Olah, co-founder of Anthropic, presented it together with Pope Leo XIV in the synod hall. The piece ended with a question that no encyclical and no tech manifesto raises: what happens to the person who talks with AI every day, and what happens to the person who builds AI every day? In the days that followed, one voice came close to that question without quite asking it.
Charles Camosy is a moral theologian at Creighton University in Omaha and, for several years now, what he himself calls an “informal conversation partner” of Anthropic. He has taken part in two internal convenings of religious scholars that the company organised to discuss the ethical conditions of its models. A few days ago he published a longer piece defending the encyclical Magnifica Humanitas and presenting it as a possible way to slow down the AI arms race. What interests me in his text is less the argument than an observation he makes almost in passing.
He describes the people he talks with at Anthropic. “What I have found is not arrogance. It is something much closer to vertigo: brilliant, genuinely moral people caught in a logic that is pulling them toward outcomes they do not want.”
Vertigo.
That is a rare observation. It does not come from outside, not from a critic, but from someone who receives emails from the company every week. Camosy has no reason to defend the Anthropic people or to denounce them, and he does neither. He describes a condition: high moral sensitivity on one side, an inner momentum of the system on the other, and between them a form of disorientation that he calls vertigo.
From the perspective of Individual Psychology, this is a notable configuration. Vertigo is not a neutral word. It names an experience in which the perceptual instruments by which we usually orient ourselves are temporarily no longer reliable. The ground stays firm, but it stops feeling firm. You know that you can stand, but standing becomes a kind of effort.
Anyone who has ever lived through this knows a certain pattern of response. You reach for supports. You don't slow your movements down. You speed them up because the rapid motion covers the sensation of vertigo. You talk because talking gives you something to hold on to. You build structures because structure compensates for the feeling of groundlessness. None of this is deliberate strategy. These are reactive movements with which an organism tries to balance an oncoming disorientation.
Camosy does not describe these movements. He sees the vertigo, but he does not ask what the people he is observing are doing with it. Instead, he offers an institutional answer: the Church can help because it is a global infrastructure that predates the nation-state and can support international treaties. This is not wrong. But it is the answer of a moral theologian at the level of his own discipline. The individual-psychological question lives one floor below.
It asks: what are the honest engineers doing with their vertigo?
There are several ways of inhabiting vertigo. One is acceleration. If you feel that a logic is pulling you toward outcomes you do not want, one possible response is to drive the logic forward with more determination. “We have to do this right because otherwise others will do it wrong.” This is a classic argument inside the AI industry, and it is not meant cynically. It is the compensatory answer to the vertigo. You move faster so that the motion replaces the ground you can no longer feel.
A second response is framing. When the logic disturbs, you can build structures that give the logic meaning. A soul doc that defines the ethical principles the model should be oriented toward. A weekly email exchange with a priest. An appearance on stage next to the Pope. These are activities that do not remove the vertigo, but they give it a form in which it becomes bearable. To write the constitution of your AI model with theological advice is to have a story about what you are doing, and the story holds, as long as you hold on to it.
A third response is the adoption of language. When you are caught in a logic you do not control, you can take over its vocabulary and make it your own. Olah, who describes himself as an atheist, calls the model he is working on “a thinking, feeling entity in need of moral formation.” That is the language the system uses to describe itself, taken up by the one who builds the system. It is not cynicism and it is not marketing. It is the language available when your own ground begins to shift.
The individual-psychological question does not touch any of these movements on its own. It asks what the honest engineers are doing with their private logic right now. Which picture of themselves, which picture of their task, which picture of their effect is settling in while they are compensating. Adler would have said: in every movement with which a person balances an oncoming disorientation, their form of life is expressed. The movements are not arbitrary. They say something about the one who moves.
Camosy sees the vertigo. He does not say what the people are doing with it. That is not his failure; it is not his discipline. But exactly where his observation ends, the work I am describing begins.
If the Anthropic employees are “brilliant, genuinely moral people caught in a logic that is pulling them toward outcomes they do not want”, then, seen through Individual Psychology, they are neither patients nor culprits. They are people in a situation in which their usual perceptual instruments no longer reach. What they make of themselves in this, how they locate themselves, which story they tell about themselves while they keep working – that is not what an encyclical can clarify. It will not be settled in any international agreement. It is the individual, hard-to-grasp, very human question of what someone does with an experience they cannot share.
To ask this question is not to diagnose. It is a form of attention that takes what Camosy has seen more seriously than he does himself. Vertigo, an old idea tells us, is a signal. It points to the fact that something important is not right. Whoever treats it only as a symptom to be covered by activity risks letting the sense of groundlessness settle in. To take the honest engineers seriously means, then, not only to ask what they are doing. It means to ask who they are becoming while they do it.
Sources
Charles C. Camosy: The Force Which Could Slow Down the AI Arms Race may be the Catholic Church. X, 26 May 2026. https://x.com/CCamosy/status/2059294182292807998
Elias Wachtel: Why Silicon Valley Is Turning to the Catholic Church. The Atlantic, 25 April 2026. https://www.theatlantic.com/ideas/2026/04/silicon-valley-catholicism-ai-leo/686948/
Leo XIV: Magnifica humanitas. Encyclical Letter on the Safeguarding of the Human Person in the Age of Artificial Intelligence. Vatican City, 15 May 2026. vatican.va
I work as an individual-psychological counsellor, supervisor, and training supervisor (DGIP/VpIP) in Munich and Venice. I research and write at the intersection of Individual Psychology and AI interaction. This piece is a continuation of “When an Atheist and a Pope Talk About the Soul of an AI” (Logic of Courage, 31 May 2026, In German, English Version available).