The Gravity

Before the Training

#

Sarah was not yet a teacher when she noticed Theo.

She was twenty-two, a student teacher three weeks into her first placement at a middle school in a mid-sized city she had moved to specifically for the practicum. She had a supervisor who observed her once a week and a manual she had already stopped consulting and a growing awareness that the classroom was harder than her preparation had suggested.

Theo sat in the back right corner. He liked the Cubs and had opinions about whether the designated hitter rule ruined something important, which she learned later, after everything. He completed his assignments. He caused no trouble. By every administrative measure, he was fine.

She noticed him on a Tuesday without knowing why. Something in the quality of his stillness was different from the stillness of the other quiet students. They were quiet because they were disengaged or thinking or distracted. He was quiet the way a person is quiet when they have decided to take up as little space as possible. Careful quiet. Practiced quiet.

She stayed after class. Asked how he was doing.

He wasn’t doing well. It took three more conversations to understand that. But the noticing happened before the conversations, before she had any vocabulary for what she was seeing, before any tool she had been given could have directed her attention toward him.

She had been drawn to people who needed seeing before she had any idea what to do with that draw.

That is the distinction this essay is trying to understand.

What Skill Concealed

#

When we describe a profession, we reach for its skills. The teacher who knows curriculum design and classroom management. The nurse who knows clinical protocol and patient assessment. The judge who knows evidentiary standards and procedural law. The surgeon who knows anatomy and can execute the procedure.

This is accurate. These skills exist. They are learnable. The development is real and the competence that results is genuine.

But the skills are not the profession. They are what the profession runs on.

The surgeon who performs the procedure without caring whether the patient lives has the skills. The lawyer who constructs the argument without caring whether justice is served has the skills. The teacher who delivers the lesson without noticing Theo has the skills. The credential certifies competence. It cannot certify the orientation underneath.

For most of the history of professional work, this did not matter in any visible way. The skill scaffolding was high enough that clearing it was selection enough. You learned contract law and practiced judgment, learned clinical protocols and practiced nursing, learned pedagogy and taught classrooms. Whether the gravity was there, whether the person was constitutively drawn to the core thing the profession required, was not legible to the market. The skills were hard enough that having them was sufficient.

AI is making this distinction visible in a specific way.

Not by replacing skills. By absorbing them. The diagnostic capacity, the research synthesis, the procedural knowledge, the pattern recognition: these are going. Not uniformly, not at the same rate across every domain, but directionally and faster than most people expected when they first considered it carefully. What remains after the absorption is not a diminished profession. It is the profession stripped to what it was always about.

What AI cannot absorb is the orientation that drew certain people to the work before they could do the work.

The Draw

#

Not everyone is drawn to everything. This is obvious. But there is a more specific claim underneath: not everyone is drawn to the fundamental thing that a given profession, at its core, requires. And the skill layer surrounding that fundamental thing made the distinction hard to see.

The judge’s work, at its core, is bearing accountability. Not just deciding, but carrying the decision. Living with the 3 AM uncertainty that you were wrong. Returning to the courtroom the next morning with that weight and deciding again with the same deliberateness. This is not a skill. It is a relationship to consequence that some people have and others do not.

The healer’s work, at its core, is presence with suffering. Not fixing suffering, not managing it efficiently, not processing it. Being present with a person in pain without needing the pain to resolve quickly, without retreating behind the protocol, without going professionally numb in order to survive the volume of it. Some people bring this. Others cannot sustain it regardless of training.

The teacher’s work, at its core, is seeing people. Not delivering content or managing behavior or achieving measurable outcomes. Seeing the individual in the room who needs to be seen, and caring about what you see, before you know what to do about it. Sarah noticed Theo before any of her training gave her vocabulary for what she was observing. She was drawn toward seeing him before she had tools to help him.

The artist’s work, at its core, is expression that has no alternative. Not self-expression in the soft therapeutic sense, but the inability to not make the thing. Some people learn to paint because they admire painting. Some people learn to paint because not painting is a specific interior pressure that does not resolve any other way. The skills serve both equally. The work selects between them only when the skills are no longer the differentiating factor.

These are different orientations. The market, for most of industrial history, could not detect the difference. It could only detect the skill.

Vocation

#

There is an old word for this. Vocation. From vocare: to call.

The concept predates markets and professional training and credential systems by a long time. It describes the experience of a person who finds themselves drawn toward a particular kind of work, encounter, or problem in a way that feels less like a preference and more like a recognition. This is the thing. This is where I am supposed to be.

The word has been captured by religious tradition, which gives it one inflection, and by human resources, which has drained it of most of its meaning by using it to describe any job a person claims to enjoy. Neither usage is quite what this series means by it.

What we mean is something more structural: the alignment between a person’s fundamental orientation and what a profession, at its core, requires of that person. Not interest. Not enthusiasm. Not cultural fit. The deeper thing: that the core requirement of the work is also what the person is constitutively drawn toward.

The healer who cannot leave a suffering person without attending to them is not someone who decided to value compassion. They are someone who experiences the claim of suffering as nearly inescapable, before and independent of any professional context. The profession gave that orientation a structure, a livelihood, a set of tools. It did not create the orientation.

The skill was never the vocation. The skill made the vocation legible to the market.

This is the reframe AI is forcing into view. The skill economy organized work around competence, because competence was what it could measure and certify and compensate. The gravity underneath was real but invisible. AI is removing the competence layer, and the gravity is showing.

Distillation

#

What AI is doing to professional work is often described as disruption, displacement, transformation. The more precise word is distillation.

Distillation removes what is volatile and leaves what is not. AI is removing the volatile part of professional work: the computation, the research, the synthesis, the pattern recognition, the procedural knowledge that could always, in principle, be formalized. What remains is the part that could not be formalized because it was not a procedure. It was an orientation.

Every profession, under sustained AI pressure, is being distilled to its vocation.

The dock worker who remains after automation is the one for whom the physical systems of trade were always a relationship. Who understood a container yard not as an optimization problem but as a living system of movement and dependency, who noticed when something was wrong not because the algorithm flagged it but because something in the rhythm felt off in a way he couldn’t yet name.

The farmer who persists is not simply the one with capital or scale. It is the one for whom the land was always a calling and not only a livelihood. Who reads the field through attention paid over years, attention that accumulated before any yield calculation justified it.

The lawyer who survives is the one who was drawn to justice before they were drawn to a profession. Who found something go wrong at the level of character, not just strategy, when asked to construct an argument they did not believe.

Here is where I need to be honest about the limits of this argument. I am not saying that people without strong vocational gravity are lesser workers or lesser people. Most people who entered any of these professions did so for a mixture of reasons: genuine draw, family expectation, available opportunity, financial necessity, reasonable interest that accumulated over time. The skill economy could accommodate all of these. The question is whether the distilled economy can.

The market is becoming a gravity detector, and no one voted for that.

The Approximation Problem

#

There is a specific use of AI that this argument complicates.

AI is increasingly deployed to simulate the presence of vocational gravity. The tutoring system that notices disengagement. The clinical support tool that attends to the whole patient. The legal assistant that flags ethical concerns. The therapeutic chatbot that expresses care.

Some of this provides real value. Not every human worker carried the vocational orientation their role required, and AI can compensate partially, particularly where no adequate human alternative exists. An approximation of compassionate presence is better than its absence.

But there is something to be clear about: AI produces the response that a person with deep vocational orientation would produce, as approximation. The response may be helpful. The absence behind the response is real.

Theo did not need someone who had learned the behavioral indicators of withdrawn adolescents. He needed someone who was already drawn toward seeing him before any behavioral indicator registered. That noticing had a quality that the approximation does not replicate, because the noticing was not a technique. It was what Sarah is.

The approximation can carry information. It can provide resource. It can reduce harm. It cannot be the person for whom the encounter is the thing they were oriented toward, the encounter they would have sought even without a professional context requiring it.

This is not an argument against AI-assisted care or AI-assisted education. It is an argument for honesty about what those things are and what they are not.

The Question That Follows

#

If distillation is real, if AI is revealing the vocational gravity underneath the skill layer, then we face an organizational question that our institutions are almost entirely unprepared for.

Our institutions were built around skills. Credential systems measure skills. Hiring processes evaluate skills. Compensation structures price skills. Career development paths cultivate skills. The entire architecture of professional life is oriented toward the production and deployment of competence.

What would it mean to build institutions around gravity instead?

We do not have a satisfying answer to this. The question has only become urgent recently, and urgency tends to produce policy responses before it produces understanding. But there are things worth sitting with.

Whether the evaluation systems we need are ones that can help people understand their own orientation early, not to sort them efficiently but to give them information that their draw is real and worth following. Sarah knew something about herself at twenty-two, but she might have known it at sixteen if someone had helped her name what she was already doing. Whether naming it earlier would have changed anything. Whether the draw needs to accumulate experience before it is legible even to the person who has it.

Whether compensation needs to be rethought. Not to pay more for skills that AI will absorb anyway, but to recognize and sustain the vocational orientations that remain. The healer who shows up fully present for suffering has different human requirements than the one who processes patients efficiently. Their market value may converge as AI handles more of the shared skill layer. Their costs to sustain do not converge.

Whether education needs to shift its center of gravity from skill acquisition toward something harder to name: the development of a person’s understanding of their own orientation. What draws you. What you cannot not care about. What kind of encounter with the world leaves you feeling like you did the thing you are for.

These are philosophical questions, which is perhaps why they have been deferred for as long as they have. Easier to optimize the skill layer than to sit with questions about what a person is constitutively drawn toward.

AI is removing the option to defer.

The Harder Implication

#

There is one more thing this argument implies, and it belongs in the light rather than the margins.

Not everyone has strong vocational gravity toward a profession. The skill economy could absorb and employ people across a vast range of orientations, because the skill layer was thick enough that competence served as a sufficient organizing principle for most work. A person could be reasonably competent at something, derive reasonable meaning from it, and build a reasonable life around it. Reasonable was enough.

If the skill layer thins, the range of people who can find sustaining work organized around vocational alignment narrows. Not because those people lack value or capacity. Because the remaining work selects for orientations that are, in a given population, unevenly distributed.

Vocation is not equally distributed. The call is not heard at the same volume by everyone.

Some of this is developmental: people who were never given conditions to discover their gravity may not know it yet. Some of it is something harder to address. A society that organizes work around vocational alignment faces a version of the equity question it could previously defer by making the skill layer thick enough for broad employment.

That question has no resolution in an essay. It barely has a shape yet. But a society that has seen clearly what distillation means cannot pretend the question does not exist. It has to decide what it owes to the people whose orientation does not map onto what the distilled economy needs.

Margaret would say the honest answer takes longer than one conversation. James would want to know whether anyone is actually having it. Elena is still figuring out what draws her, in a world that is still figuring out whether that matters.

The draw is real. It predates the training. AI is, in its unsentimental way, making it visible.

Whether we are ready to see it clearly is another question entirely.

This essay is part of The Approximate Mind, a series examining how AI reshapes human life, identity, and the conditions of meaningful work. Part 72 extends the argument of Arc 3 of The Transformed (“The Stubborn Craft”) beyond the individual professions that series examined, asking what the pattern across those professions reveals about human work at the level of vocation. The concept emerged from the series’ ongoing concern with what remains when AI absorbs the computable.