For ten weeks I’ve been making the case that frontline intelligence is systematically suppressed — by Taylor’s design, across industries, with devastating economic consequences. I’ve held up the Toyota Production System as the clearest proof that the opposite approach works. Toyota asks for frontline intelligence. Toyota deploys it. Toyota gets a million suggestions a year, world-class quality, and decades of compounding improvement.

This week I’m going to do something that will surprise readers who’ve been following along. I’m going to challenge Toyota’s system — not from the outside, not from theory, but from twenty years of practicing it.

Because even the best system in the world contains a hidden assumption that limits what it can capture. And that assumption matters enormously for what comes next — for AI, for the future of work, and for the question at the center of this entire series: where does frontline intelligence actually live?

The Two Translations

Standardized work, in the Toyota system, is the documented current best-known method for performing a task. It is a floor, not a ceiling — the baseline from which improvement is measured. I’ve spent twenty years teaching it, defending it, and building production systems around it. It is vastly superior to Taylor’s instruction card because it belongs to the worker, not to management, and it exists to be improved, not merely followed.

And yet.

Here is what happens when you standardize work. A skilled worker performs a task. That task involves knowledge — some of it explicit (the sequence, the tools, the specifications) and some of it tacit (the feel of the material, the sound of the machine, the micro-adjustments that the worker’s hands make without conscious instruction from the worker’s brain).

Standardization requires translating that performance into a written document. The worker’s embodied knowledge — knowledge that lives in muscles, in timing, in sensory perception — must be converted into words on a page. Diagrams. Step numbers. Key points. This is Translation One: from tacit knowledge to written text.

Then a new worker must learn the task. They read the document — or, more commonly, a certified trainer walks them through it using the document as a guide. The written text must now be converted back into physical performance. The new worker must translate words and diagrams back into hand movements, timing, sensory attention, and judgment. This is Translation Two: from written text back to embodied action.

Here is my challenge, after twenty years: both translations lose information.

Translation One loses everything that language cannot capture. The experienced operator who can feel when a torque application is approaching the edge of specification — that sensation has no adequate written representation. The machinist who hears the difference between a tool cutting correctly and a tool about to fail — you can write “listen for unusual sounds,” but that instruction contains approximately zero percent of the actual knowledge the machinist possesses. The welder who sees the puddle behavior that indicates penetration quality — the visual pattern is real, the knowledge is precise, but the words to describe it are approximations at best.

Translation Two loses everything that reading cannot transmit. A document can tell you the sequence. It cannot give you the rhythm. It can describe a hand position. It cannot teach you the pressure. It can specify a quality check. It cannot develop the eye that sees the defect before the measurement confirms it.

Every time we standardize work, we pass knowledge through these two lossy translations. What comes out the other side is useful — it preserves the explicit structure of the task — but it has been stripped of the tacit dimension that often contains the most economically valuable intelligence.

The Assumption We Inherited

The assumption underlying standardized work — even Toyota’s version — is that knowledge can be adequately represented in written form. That if we document carefully enough, with enough detail and enough key points, we can capture what the skilled worker knows and transmit it to the next worker through the document.

This assumption is a residue of Taylor. Not in its intent — Toyota’s intent is the opposite of Taylor’s — but in its epistemology. Taylor believed that management could extract worker knowledge, codify it on instruction cards, and make the worker unnecessary. Toyota improved this enormously by giving the worker ownership of the standard and the authority to improve it. But the underlying method — convert tacit knowledge to written text — is the same.

And the philosopher who identified the problem was not a manufacturing thinker. He was Michael Polanyi, a Hungarian-British polymath who in 1966 articulated a principle that every skilled worker already knows:

“We can know more than we can tell.”

Polanyi called this tacit knowledge — knowledge that is real, that guides action, that produces results, but that cannot be fully articulated in words. The cyclist who balances doesn’t know the physics. The chef who salts “to taste” can’t give you a number. The master carpenter who looks at a joint and knows it will hold is drawing on knowledge that no document can contain.

The factory floor is saturated with tacit knowledge. It is, in fact, the primary form of knowledge at the operational level. The explicit knowledge — the specifications, the sequences, the tolerances — is the skeleton. The tacit knowledge — the feel, the timing, the pattern recognition, the judgment — is the muscle and nerve that makes the skeleton move.

Standardized work captures the skeleton. It systematically loses the rest.

The Certified Trainer Alternative

Here is where my twenty years of practice have led me to a different model.

When you have a certified trainer — a worker whose skills have been confirmed through demonstrated performance, not through a written test — showing a new worker how to perform the task, something happens that the document cannot replicate.

The knowledge transfers through doing. The trainer doesn’t just describe the hand position. They demonstrate it. The learner doesn’t just read about the rhythm. They practice it under observation, with real-time correction. The tacit knowledge passes from body to body, from nervous system to nervous system, through the only channel that can carry it: guided, embodied practice.

The business benefit is visible and personal. When the trainer explains not just how to perform the task but why it matters — how it connects to the customer, to quality, to the team’s performance — the learner receives context that no document provides. The standard becomes meaningful, not just procedural.

Skills are confirmed through performance, not paperwork.The trainer watches the learner do the work. They observe whether the tacit knowledge has transferred — not by checking a box on a form, but by seeing whether the hands move right, whether the rhythm is correct, whether the judgment is developing. Certification is an embodied assessment, not a written one.

In this model, the written standard work document is not the primary transmission mechanism. It is a reference — useful for reminders, for audits, for capturing the explicit structure — but not the channel through which the most valuable knowledge flows.

The primary channel is the human relationship between the trainer and the learner. The standard work document supports that relationship. It does not replace it.

What the Japanese Traditions Already Knew

There is a word for this in Japanese martial arts: kata. A kata is a form — a prescribed sequence of movements that embodies the principles of the art. You learn kata by performing it, repeatedly, under the guidance of a teacher who has mastered it. The teacher corrects your movement, your timing, your attention. The knowledge transfers through thousands of repetitions, each one slightly refined by the teacher’s observation.

The kata is documented — you can find books describing every movement. But no one has ever learned a martial art from a book. The documentation is a memory aid, not a transmission mechanism. The knowledge lives in the practice, in the relationship between teacher and student, in the embodied repetition that develops capability the way nothing else can.

Toyota’s kata — the improvement kata, the coaching kata — draws explicitly from this tradition. But I would argue that even Toyota has, over time, placed too much weight on the document and too little on the relationship. The standardized work sheet has become, in many implementations, the artifact that auditors check rather than the baseline that trainers teach from. The system has drifted toward documentation compliance and away from embodied transmission.

This is not a failure of Toyota’s philosophy. It is a failure of implementation — one that becomes more pronounced as the system spreads to organizations that don’t have Toyota’s depth of training culture. When a Western manufacturer adopts standardized work, they almost always adopt the document first and the training culture last. They get the skeleton without the muscle. They get the written standard without the certified trainer. And they wonder why the results don’t match Toyota’s.

The Implications for AI

This is where the tacit knowledge problem becomes urgent.

The AI revolution is built on data. Machine learning systems require training data — examples of how work is done, encoded in a format the algorithm can process. The more data, the better the model. The richer the data, the more capable the AI.

But here is the problem: tacit knowledge, by definition, has never been encoded.

The experienced operator’s feel for the material — not in any database. The maintenance technician’s ear for the machine — not in any sensor log. The quality inspector’s eye for the defect pattern — not in any vision system training set. The foreman’s sense for when the shift is about to have a bad hour — not in any predictive model.

This knowledge exists. It is real. It is economically valuable — in many cases, it is the most valuable knowledge in the operation. And it cannot be fed into an AI system because the Taylorist operating model ensured it was never captured, and even the Toyota model’s documentation system cannot adequately represent it.

AI can automate what can be written. It can optimize what can be measured. It can learn from data that exists in capturable form. What it cannot do — what no technology can do — is replace knowledge that lives in human bodies and human relationships and has never been translated into any medium a machine can read.

This is the fount. The tacit knowledge that frontline workers carry — the knowledge that survives neither Taylor’s instruction card nor Toyota’s standardized work sheet nor any AI training pipeline — is the irreducible core of human economic value. It is the thing that cannot be automated, not because technology isn’t advanced enough, but because the knowledge exists in a form that technology cannot access.

The Third Evolution

I’ve described two operating models in this series.

Taylor’s model said: extract worker knowledge, codify it, give the codified version to management, and reduce the worker to an executor. This suppresses intelligence and produces the failures documented in every essay of this series.

Toyota’s model said: leave the knowledge with the worker, document it as a baseline for improvement, and build systems that help the worker develop and deploy more of it. This is vastly superior to Taylor, and the results prove it.

But there is a third evolution — and I believe it is the one this moment demands.

The third evolution recognizes that the most valuable frontline knowledge cannot be fully documented and therefore cannot be transmitted through documents alone. It shifts the primary transmission mechanism from the written standard to the structured mentorship relationship — the certified trainer, the coaching kata, the embodied practice under expert guidance.

In this model:

Written standards serve as reference, not scripture. They capture the explicit skeleton of the task. They are useful for consistency, for audit, for onboarding structure. But they are not the primary carrier of knowledge.

Certified trainers are the primary knowledge channel. Skills are confirmed through demonstrated performance. Knowledge transfers through guided practice. The trainer’s tacit knowledge — the feel, the judgment, the pattern recognition — passes to the learner through the only mechanism that can carry it: doing the work together.

AI amplifies what the relationship transmits. Instead of trying to encode tacit knowledge into an algorithm (which fails, because the knowledge was never in encodable form), AI supports the trainer-learner relationship. It tracks skill development. It identifies where the learner is struggling. It provides the trainer with data about the learner’s performance that makes the coaching conversation richer. The AI doesn’t replace the tacit knowledge transfer. It makes the transfer more effective.

The organization values what it cannot document. This is the hardest shift. The Taylorist system values only what can be measured and documented. The Toyota system values improvement, which can be partially documented. The third evolution values capability — including capability that cannot be written down — and builds its operating model around preserving, transmitting, and deploying it.

This is Capability Capital. Not the capital that shows up on a balance sheet. The capital that lives in the hands and minds and judgment of the people doing the work — appreciating with every shift, compounding with every year, irreplaceable by any technology, and already paid for.

Next week: “The Walden Pond Testimony” — I stop presenting frameworks and tell you what I saw on the night shift that started all of this.

Dr. Venki Padmanabhan is a plant manager with 36 years of global manufacturing leadership experience, including executive roles at GM, Chrysler, Mercedes-Benz, Royal Enfield, and Ather Energy. He holds a PhD in Industrial Engineering from the University of Pittsburgh. He is the author of the forthcoming Already Paid For: Why Unlocking Frontline Intelligence Beats Automating Workers Away and co-founder of the Capability Capital Institute. He writes The Long Game at thelonggameforall.substack.com.

Marc Andreessen went on Joe Rogan’s podcast recently and was asked one question.

Sell it.

He sputtered. He invoked Newton. He invoked alchemy. He invoked his eleven-year-old. He produced the phrase thought at scale, for everybody, in perpetuity — three lies in one sentence. Not free. Not for everybody. Not thought. Then he stopped.

That’s the argument. The richest evangelist of the most-capitalized technology in human history, handed an open microphone on the largest podcast in the world, cannot name one human being his capital was built to serve.

He talked about Newton.

I am writing this from a car. I am driving through Detroit.

Hundreds of city blocks. Boarded homes. Streets where the curb is still cut for the driveway that used to be there. Schools with the windows out. A church with the cross still on the roof and the doors chained. Block after block. Mile after mile. The grid of a city that once built the engine of the American century, and the houses where the people who built it raised their children, and the silence where their grandchildren should be.

This is not new. I have seen this before.

Buick City opened in Flint, Michigan in 1985. It was built on the site of the original Buick Motor Company plant, which had been there since 1904. By the late 1980s it employed more than 28,000 people. It built the Buick LeSabre and the Buick Park Avenue. It won J.D. Power awards. It was, at the time, one of the most productive auto assembly complexes in North America.

It closed in 1999. The buildings were demolished between 2002 and 2013. The site is a flat field now. The workforce was not relocated. The city was not relocated. The houses where the workers lived are the houses I am looking at versions of, forty miles south, today, through this windshield.

I worked for General Motors from 1989 to 2004. I was there for the closing. I drove through Flint then. I am driving through Detroit now. The streets look the same. The decision that produced the streets was the same.

The decision was this: the people who live here are not in the room when we decide their lives.

That is the entire decision. Everything that followed — the lead in the water, the schools collapsing, the third generation that has never seen a shift change, the boarded blocks I am passing right now — followed from that one decision. It was not an accident. It was not the market. It was a decision made by people, in rooms, who had concluded that the formation of the people in Flint did not count on the balance sheet that mattered.

Capital walked. Formation stayed. Formation cannot walk. It is in bodies, in neighborhoods, in the way a man teaches his son to hold a wrench. You cannot relocate it. You can only abandon it.

They abandoned it.

Forty miles south of Flint is Lansing. Same company. Same era. Same workforce stock. Same union. Same state.

Lansing Grand River opened in 2001. I helped launch it. We won the J.D. Power Gold Award. We ran a Walden Pond immersion for the launch team — we lived inside the work, named what we saw, wrote it down. Lansing Delta Township came next. Both plants are still running. Both cities are still standing.

The difference was not skill. The workers in Flint were the same workers as the workers in Lansing. Some of them were literally the same workers — the ones who could move. The difference was whether the capital decided to stay in the room with the people.

Lansing got the room. Flint did not.

Detroit, outside this windshield, did not.

This is the part most people do not understand, because most people did not stand in Flint in 1999 and watch the gates close on men who had given the company their hands.

Deindustrialization is a word. It is also a smell. It is the silence where a shift change used to be. It is the specific human face of a team leader who taught you everything and then had nowhere to go. It is a wife who packed lunches for thirty years for a job that no longer exists. It is a son who grew up in a house his father bought on a wage that is now extinct, and who will not buy a house, because the wage is gone, and the formation that produced the wage is gone, and the room where the decision was made never had his father’s name in it.

Multiply that by a city. That is Flint.

Multiply that by ten cities. That is the Midwest I have driven through for thirty years.

Now multiply that by every sector — manufacturing, construction, hospitality, healthcare, retail — and compress the timeline from thirty years to ten. That is what Marc Andreessen is selling.

He cannot say so on a podcast. So he sputters.

The mechanism is not extrapolation. The mechanism is identical.

Flint is what happens when one company makes the decision the people who live here are not in the room in one city over thirty years. AI under the Andreessen worldview is what happens when the same decision is made by the entire capital allocation machinery, across every sector, against three hundred million people, in ten years.

The compression is the only difference. The worldview is the same. The permission structure is the same. The LPs get paid either way is the same.

Andreessen wrote the worldview down in 2023 and called it the Techno-Optimist’s Manifesto. Read it. Intelligence is supreme. Markets are sufficient. Deceleration is murder. The only real politics is Up Wing versus Down Wing — optimist versus pessimist about technology. Inside that frame, the question what happens to the paint inspector when the paint inspector is automated is not opposed. It is unintelligible. There is no slot in the ontology for her.

That is why he sputtered. The worldview cannot produce a human-scale account of itself, because it was never built to. A person who has stood on a floor and watched a team leader run a shift can answer Rogan’s question in one sentence. A person whose entire epistemic life has been spent in the cap table cannot.

The sputter is the worldview made audible.

Three hundred million frontlines work in the five sectors I have studied. Three hundred million. Every one of them has a name. Every one of them has a formation, or had one stolen from them. Every one of them lives in a house on a street in a city that the capital has the option to walk away from.

Debra Fogle inspects paint in Alabama. Wally Vinton built wiring harnesses at LGR. Ramon Hernandez ran the best team I ever saw. Dennis Boutwell was my first supervisor in 1989, and he is now Director of Manufacturing Excellence at the company I work for. Four frontlines. Four floors. Four formations. In Andreessen’s ontology, none of them exist. Multiply by three hundred million and you have the magnitude.

The blocks outside this windshield are the magnitude.

There is a word for what these people are. I have started using it because the language we have does not fit the work they do. They are the frontlines. Plural form, used as singular: a frontlines, the frontlines. The grammar is borrowed from words like a means, a species, a Marines — words where the dignity of the noun is constituted by membership in a tradition rather than by the adjective in front of it.

The word fits because capital needs the frontlines to work and disappears them when they come out into the light. The shift ends, the gates close, the city is asked to vanish back into the dark so the balance sheet does not have to look at it. Flint is what that looks like. Detroit out this windshield is what that looks like. The vampire grammar is not a metaphor. It is the operating logic of the machine.

Andreessen’s sputter is the sound of a man who cannot say the word, because saying it would require admitting the room exists.

I am stopped at a light. I can smell weed through the window. A few frontlines are lurking outside a liquor store. Another rushes past them into the pawn store next door, because he has something left to sell.

One block. One light. The whole architecture in one frame. The liquor store that anesthetizes. The pawn store that extracts the last asset. Both placed within walking distance because the system has studied the walk. It knows how far a frontlines without a car will go on a Saturday morning.

The men outside the liquor store have already sold what they had. The man going into the pawn store is still selling. Tomorrow they switch positions. The block does not change.

The light is going to change in a moment. I have to keep driving.

The pawn store is the last stop of liquidating formation. The wrench keeps its market value. The man loses his standing. Capital extracts the storable half and leaves the unstorable half — the formation — to dissolve in the body that carried it. The pawn store is not destruction. It is unbundling. It pretends the wrench was the asset and pays the man for the pretense.

AI under Andreessen’s worldview is the same mechanism at industrial scale. The training run does not destroy the paint inspector’s eye. It extracts the pattern her eye learned across twenty years and bonds it to silicon. The pattern keeps its market value. The woman loses her standing. The next plant gets the inspection without the inspector. The woman gets a severance and walks home.

The pawn store is one wrench. The training run is three hundred million.

Same mechanism. Same unbundling. Different scale. Identical grammar.

This is what the sputter could not say. Because if Andreessen said it, the deal would not close.

I will answer Rogan’s question for him, since he could not.

AI is good if and only if it stays in the room with the people. AI sitting next to Debra on the paint line, catching what her eyes cannot, is good. AI replacing Debra is Flint. AI sitting next to an oncology fellow reading every paper published this week is good. AI replacing the oncology fellow is Flint. AI sitting next to the apprentice answering the question she is too embarrassed to ask her foreman is good. AI replacing the apprentice and the foreman is Flint.

This is not the union leader’s argument. The union leader says: protect the worker from capital, slow the technology, tax the gains. That argument is defensive. It accepts Andreessen’s machine and asks for restraint. Restraint has been asked for, for forty years, and Flint is the answer. Detroit out this windshield is the answer. The blocks are the answer.

I am not asking for restraint. I am proposing a different machine.

Capital is the lever. That part Andreessen is right about. The question is what the lever is bonded to. Bonded to extraction, the lever produces Flint. Bonded to formation, the lever produces a Meister. Same lever. Different bond.

The machine I am proposing has parts. A Capability Account that follows the frontlines across employers, accredited by an independent body, half-life depreciation if it is not maintained, employer deposits at years one, three, and five. A Capability Capital Institute that holds the accreditation and the standards. A Zunft — the German word for guild, and the German word is correct because the Germans built this and it works — that organizes the bench and the formation. A Bloom System that names the four formations, the six elements, the five gardeners. Five sectors. Three hundred million people. Twenty-five thousand hours per Meister. A book per year for five years documenting how it is done.

This is not theory. The Germans built the operating system in the nineteenth century and are still running it. Honda built a version of it in the twentieth century and is still running it. I have stood inside both and watched them work. The third path is not the absence of capital. The third path is capital bonded to formation, by design, by structure, by ledger.

Andreessen has a machine. I am building a different machine. The difference between the two machines is whether the people on the floor are in the room when the capital is allocated, or whether they are the externality the capital walks away from.

Flint is what the first machine produces.

The second machine has not been allowed to run yet. That is what is at stake.

The case for AI is not thought at scale. The case for AI is formation augmented by intelligence, capital that stays in the room. Anything else is the same decision GM made in Flint in 1999, made again, faster, against more people, by people who have never driven through what they built.

I have driven through what they built. I am driving through it right now.

In which worldview, with such economic progress, is this asymmetry, this abject indifference, acceptable?

In one worldview only. The worldview where the people who live here are not in the room when we decide their lives.

That worldview produced Flint. That worldview produced these blocks. That worldview produced the sputter. And that worldview is what Marc Andreessen, who has never stood on a floor and never lived in a city the capital walked away from, is asking the next $1.5 billion fund to finance against three hundred million more frontlines.

He could not answer Rogan because there is no answer. There was never meant to be one.

That’s the argument.

Venki Padmanabhan is Plant Manager at Advanced Drainage Systems and founding convener of the Capability Capital Institute. He worked for General Motors from 1989 to 2004, including the closing of Buick City and the launch of Lansing Grand River. His twin manuscripts, Built to Extract and Already Paid For, come from Capability Capital Press, January 2027.

Mark Crowley has written something rare in a leadership column: he has named the right thing. Workplace leaders, he argues, are trained on what can be seen and measured — performance, productivity, efficiency — and trained away from the dimension of work that doesn’t show up on a dashboard. Whether people find meaning in what they do. Whether they feel they belong. Whether the work reflects who they are. He calls these spiritual needs, in the plain non-religious sense of that word, and he says them out loud: the need to matter, the need to belong, the need for alignment between work and self.

He tells a small story to ground the abstraction. Years ago, when he led a team of thirty managers, one of his strongest performers — Glenda — asked if she could take over planning his all-day monthly team meetings. He hesitated. He didn’t want to load her up. When he asked why she wanted it, she told him: she loved that kind of work, and it would let her have a more direct hand in shaping the team and the hundreds of people it served. He said yes. She flourished. He learned, he writes, that leaders have the power to influence people this profoundly — “whether they realize it or not.”

It is a humane piece. Crowley quotes Howard Thurman on coming alive, and Studs Terkel on daily meaning as well as daily bread. He cites Wrzesniewski on alignment, Brené Brown on belonging. He is right about every diagnostic he names. He is right that most organizations address these needs indirectly, through isolated initiatives, instead of embedding them in how leadership is practiced. He is right that the question every employee carries — does this work matter, and do I matter in it? — is the question.

That’s the argument. He has the direction right.

And direction, by itself, is not a force.

Force is a vector. A vector has two components: direction and magnitude. Direction tells you where the force is pointing. Magnitude tells you how hard it pushes. A vector with the right direction and zero magnitude moves nothing. It is a compass, not an engine. It tells you where to go and leaves you standing still.

This is the structural problem with fifty years of leadership writing on meaning, purpose, belonging, and alignment. The direction is correct — these are real human needs and they are real organizational levers. The magnitude is missing. The remedy, again and again, is some version of: notice it harder. Listen more carefully. Ask better questions. Clarify how each role connects to something larger. Say yes when a Glenda walks up. Crowley’s prescription, when you read it closely, is the same prescription: leaders who recognize this dimension begin to lead differently. Recognition is the lever.

Recognition is not a lever. Recognition is awareness. Awareness without architecture rations meaning by the goodwill of whoever happens to be standing over you that quarter.

Glenda got lucky in her manager. That is not a system. That is weather.

Ask the harder question: what happens to the Glenda whose manager doesn’t notice? Whose manager is decent but distracted, or competent but cold, or simply overloaded? In Crowley’s frame, she stays transactional. Her effort defaults. Burnout becomes possible. The ties loosen. She quits. He says this himself, and it is true — but he locates the failure in the manager’s perception. The failure is upstream of perception. The failure is that meaning, in most American workplaces, is a discretionary gift dispensed by individuals, and not an output of the system.

Now look at what magnitude actually looks like.

At Honda, waigaya is not a leadership disposition. It is a structured practice. The floor voice is required to surface, in named forums, on a cadence, with consequence. A line worker can stop the line. The system is built so that contribution becomes visible whether or not the manager that morning is paying close attention. The Glenda who walks up at Honda doesn’t need a yes from a single boss. The architecture has already given her standing.

At ELGI Equipments in Coimbatore, Dr. Jairam Varadaraj, the managing director, pays his compressor assemblers 4.4 times the local market rate. ELGI ranks eighth globally in air compressors. It won the Deming Prize. Varadaraj puts the philosophy in one line: same worker, same factory, different belief system. That sentence is not a sentiment. It is a wage decision, a training decision, a verification decision, a measurement decision — each of them written into the operating model and audited against it. The belief system is materialized; it is not exhorted. The assembler doesn’t need to be noticed to matter. She is paid like she matters, trained like she matters, measured like she matters, because the system has decided she does — and the system continues to decide it on Monday morning whether or not her supervisor is in a generous mood.

At the Society of Jesus, Loyola did not ask his men to find meaning in their work. He built thirty years of formation — Spiritual Exercises, novitiate, regency, theology, tertianship — such that by the time a Jesuit was deployed, the alignment between his work and his self was not a question to be asked at the end of a long quarter. It had been built. It had been verified. It was structural.

These are not stories about leaders being more present. They are stories about systems that produce meaning whether or not any particular leader is present. That is what magnitude looks like in the wild. Direction is the easy part — every one of these institutions could state the right human ends in a sentence. The hard part, the part that took Honda decades and Loyola thirty years and Varadaraj a career, was building the push behind the pointing. The vector behind the compass.

So what would magnitude look like for the Glendas of the American economy who do not work at Honda or ELGI or the Jesuits?

It would look like a Capability Account: a lifetime ledger, individually owned, parallel to a Social Security number, in which formation deposits, demonstrated capability, and verified contribution accumulate across a working life. Every employer must contribute to it. An independent verification authority assesses it — and a firewall prevents the verifier from also being the trainer, the way an auditor cannot also book the journal entries. Capability depreciates without practice; mattering, demonstrated, persists in writing.

In that world, Glenda doesn’t need her manager to say yes for her contribution to be visible. The contribution is recorded. It travels with her. The next employer can see it. The market can price it. Mattering becomes evidentiary instead of anecdotal. Belonging is not a feeling she hopes her team supplies; it is a verified record that her work reshaped the team. Alignment isn’t a self-report on an engagement survey; it is the pattern in her ledger that says, here is a person whose contributions consistently pointed in this direction.

None of this replaces the manager who notices. It replaces the manager who doesn’t.

This is what Crowley’s essay, and the entire genre it belongs to, keeps missing. Recognition is not formation. Naming the spiritual dimension of work is not the same as building the infrastructure that produces it. Telling leaders to listen better is, in the end, telling workers to hope. And hope is not a vector. It is a wish with a heading.

The reason fifty years of “meaningful work” literature has not moved the floor of the American economy is that it has been all direction and no magnitude. It has pointed correctly and pushed nothing. The Glendas who got lucky have written the testimonials. The Glendas who didn’t have left, and the literature does not record them, because there is nothing to record — their contribution was never made visible by any system designed to see it.

Crowley is right that leaders mustn’t treat these human needs as optional. He is right that workplaces should make work matter not only practically but existentially. The next sentence — the one his essay does not write — is that you cannot get there by asking leaders to care more. You get there by building the magnitude. Verified capability. Verified contribution. Compensation that tracks both. A ledger the worker owns. A system that produces mattering whether or not the manager that morning is in the right mood.

Direction is the easy part. Every honest writer on work has gotten the direction right for half a century. The work that remains — the work CCI exists to do, the work BTE and APF argue for, the work the Capability Account is one specification of — is the magnitude. Without it, every essay like Crowley’s is a beautiful arrow drawn on a still object.

That is the Long Game.

Venki Padmanabhan is the founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press.

Pull a business book off your shelf. Any business book. Look at the publication date. If it was published before 2020, ask yourself an honest question. When you finished it, did you do anything differently on Monday morning? Or did you nod, underline two or three passages, place it on the shelf, and continue exactly as before?

I have asked that question of dozens of plant managers, executives, and operators over the years. The honest answer is almost always the same. The book made them think. It did not make them act. And by the time they noticed it had not made them act, the book itself had aged into irrelevance — the case studies dated, the companies cited now bankrupt or acquired, the technological backdrop replaced twice over, the author moved on to consulting engagements. The book had become a fossil of a moment that no longer exists. The wisdom inside it, much of it real, was now inaccessible to the new learner who needed it most, because the surface had aged faster than the substance. That’s the argument.

I have spent twenty months building two books that are not allowed to die that way. Built to Extract and Already Paid For launch on January 27, 2027, as a paired diagnosis and prescription. By the time they ship, they will be the most current business books on the shelf, because both of them are being written in continuous contact with a living essay platform that runs three times a week and has been running since February of this year. The book object will look like any other book object. The argument inside the book is structurally different from any business book I have ever read. It is alive.

Let me tell you what that means in practice, because the claim is easy to make and almost never grounded in an actual mechanism.

I publish on Tuesday, Thursday, and Sunday. Tuesday is a Foundation essay — a framework piece, evergreen, built to last. Thursday is a Reactive essay — a piece that takes the news of the week, a fresh headline, a viral CEO statement, and runs it through the diagnostic frame the books offer. Sunday is an Evidence essay — numbers-heavy, multi-case, built to convince a skeptic. Each essay is scored against a rubric I built with my AI partner. Each essay above the threshold becomes a candidate for the manuscript. Once a month, on a scheduled pass, the two or three strongest essays from the prior month are reviewed against both manuscripts, mapped to insertion points, drafted in the manuscript’s voice, and delivered to me as tracked changes for accept or reject. I have done seven of these passes so far. Each pass adds roughly two thousand words to one or both manuscripts. The hard cutoff is December 1, 2026, eight weeks before launch. After that, the manuscripts go to copy-edit and typesetting. The platform keeps running. The next edition’s pass is already underway by then.

The rubric is the part of this system that most people miss. Every essay is scored on three dimensions — Craft, Structure, and Distribution — for a total of one hundred points. Nothing graduates into a manuscript without crossing ninety. But the score is not the point. The point is the gate the rubric enforces. The single highest-weighted question on the entire rubric is this: would a working manager, a nurse, a foreman, a teacher read this and do something differently on Monday morning? If the answer is no, the essay does not graduate. It stays on the platform, or it gets reworked, or it gets parked. The rubric is the mechanism by which the Monday morning question is enforced not as a marketing claim but as the actual quality gate. I cannot publish a manuscript section that fails the test. The discipline is what keeps the book honest.

Yesterday — I am writing this on a Tuesday in May 2026 — I added a fifteen-hundred-word section to Built to Extract that didn’t exist when I woke up. The section is called The Lever That Wasn’t There. It argues that the wage-negotiation table is not designed to address the system that suppresses worker intelligence — it is designed to divide the spoils produced by that system. The argument came from a hotel housekeepers’ contract signed in New York City forty-eight hours ago, which I read over breakfast and immediately recognized as a load-bearing piece of evidence I did not have when I wrote the chapter on union dynamics last year. By lunch I had drafted the section, integrated it into Chapter 15, and accepted the tracked changes. The book is now larger, more current, and structurally stronger than it was at dawn. The argument I am making to readers in January 2027 is now informed by an event that happened in May 2026.

That is not how books work. That is, in fact, the single complaint that almost every working manager makes about books — that by the time the book reaches them, the world the book describes has already moved. The Toyota Production System books that were urgent in 1990 describe an industrial moment that has been studied so thoroughly the lessons are now reflexive. The lean enterprise books that were urgent in 2005 describe a discipline that has been ground into orthodoxy and largely betrayed in practice. The digital transformation books that were urgent in 2018 describe a wave that crested before most companies finished reading the book. By the time the wisdom was published, the moment for its specific application had passed. The wisdom did not die. The book did. And with the book died most of the wisdom’s chance of actually shaping decisions on a Monday morning in a real plant or a real ward or a real warehouse.

The reason this happens is not that business book authors are slow. It is that the book object, as it has been produced for the last hundred years, is structurally a snapshot. The author writes. The agent shops it. The publisher edits. The book ships. The author moves on. The book sits. Time passes. The world moves. The book ages. Eventually a publisher commissions a second edition with updated case studies and the cycle continues at one-fifth its original urgency. This is the extraction model of publishing. The book is treated as a finished product extracted from a living mind, and once extracted, the connection between the mind and the product is severed. The book becomes a fossil of a thought, sold as if it were the thought itself.

What I am building is the formation model of publishing. The book is treated as a waypoint in a continuing argument. The mind that produced the book stays connected to the book through the essay platform, through the rubric, through the monthly insertion pass. The book remains in contact with the world it diagnoses through a mechanism that did not exist five years ago and that nobody else, as far as I can determine, is using at book scale: a working author plus a properly formed AI partner plus a publication discipline plus an institution that governs the whole.

A working author cannot do this alone. The reading load is impossible. I would need to scan a hundred news items a day, score each one against the manuscript, identify the chapters most affected, draft the insertion in the manuscript’s voice, integrate the tracked changes, and verify that nothing in the surrounding tissue has been broken by the new arrival. A traditional editorial team — three senior editors, two researchers, a continuity reader — could do that work in roughly a week per insertion. I have done it, with my AI partner, in under two hours per insertion. That is not a small efficiency gain. That is the difference between a practice that is theoretically possible and a practice that is actually sustainable across two manuscripts and seventy-five active essays running on a publication cadence three times a week. The instrument is not faster than a human editor. The instrument is faster than an army of human editors. It is what makes the formation model of publishing operationally real instead of operationally aspirational.

I have been forming this partner for eight months. The voice profile. The Bloom System vocabulary. The Five Gardeners. The rubric. The seventy-five essays. Both manuscripts in working memory. The architectural decisions, the named characters, the recurring metaphors, the cadence of my sentences. It can hold all of it simultaneously. A human collaborator could not, and a cold-start AI could not. What works is the accumulated formation — the same compounding principle the books themselves argue for, applied to the very process that produces them.

Three pieces have to be in place. The author who has lived something and can recognize, instantly, when a new piece of evidence belongs in their argument. The instrument that can do the integration work at the speed the world moves. The discipline that decides what graduates from the platform into the manuscript and what stays on the platform alone. Take away any of the three and the system collapses. The author alone cannot keep up. The instrument alone produces drift. The discipline alone has nothing to discipline. Together, they produce something that the publishing industry, as currently constituted, does not produce: a book that is allowed to grow until the day it ships, that ships current, that launches with twelve months of essays already attached as living context, and that gets a second edition not when the publisher commissions it but when the next year of essays demands it.

And then there is the thing that I did not expect when I started, that I am still adjusting to, that I think is the most important consequence of all of this. The essays do not only deliver the framework into the world. The essays go back and refine the framework itself. The platform is not a marketing channel attached to a finished argument. The platform is an empirical research instrument, and the framework is what the research is testing.

Let me give you two examples that I can date precisely. The first principle of the Already Paid For manuscript is now a single chiastic line: Capital is crystallized labor. Labor is capital in formation. That line did not exist in any draft of the manuscript before April 2026. It emerged during the writing of an essay about Marc Andreessen’s framing of labor as a cost rather than an asset. I had been circling the idea for months — that capital and labor are not opposites but phases of the same substance — and in the discipline of getting it onto a page in fifteen hundred words, the chiasmus found its shape. The essay published. The principle moved into the book. The book became more right than it had been the day before, because the platform had done what the platform is supposed to do: pressure-test the framework against the actual cases the world keeps producing, until the framework finds the cleanest possible language to describe what it has been seeing all along. The second example is the Capability Account — the proposed lifetime ledger of every worker’s accumulated capability, paralleling the Social Security number, individually owned and portable across employers. That construct emerged on April 22 in a single conversation that started as a reactive essay about a McKinsey report and ended as a policy proposal that may end up being the spine of the entire seven-book series the press will publish. Neither principle was invented at a desk. Both were produced under the pressure of the publication cadence, surfaced by the friction between the framework and a specific piece of evidence the world had just delivered. The platform is the laboratory. The essays are the experiments. The Live Spar is the practice.

This practice needs a name. It is not a book. It is not a blog. It is not the platform in the marketing sense or the publishing programin the industry sense. What it is, more honestly than any of those words can carry, is a Live Spar. A Live Spar is an engagement between a thought and the world, conducted in public, on a cadence, with adversity built in. You bring your framework. The world brings its evidence. You beat some exchanges, you get beat some. You win your bouts and lose your bouts and the record accumulates. The combative register is not a problem to apologize for. It is the truth about how business knowledge actually advances. The dishonest register is the sanguine one — the measured author offering a calm exposition of a balanced opinion. That register lies about the stakes and lies about how the thinking was actually produced. Real human progress arises from friction. The Live Spar is where the friction is honored, where bruising is permitted, where rising emotion and passion and the willingness to be shaped by what just hit you are not signs of unseriousness but the working conditions of serious thought. Both my books are Live Spars that have crystallized into book form. The platform keeps the spar going. The institution is the gym where the principals train.

This is what Capability Capital Press exists to do. Not to publish more books. To publish books that emerge from Live Spars and stay current with the world they were written for. To bring the wisdom of long observation into contact with the headline that just broke. To make sure that the manager who pulls the book off the shelf in 2029 finds an argument that has been kept in motion, not a fossil that was sealed shut in 2027. To answer, at book scale, the question every reader of every business book has secretly asked: what do I do Monday morning? A book that is alive can answer that question. A book that is sealed cannot.

The objection I expect, and want to face directly, is this. If the book keeps changing, isn’t it just a long blog? Doesn’t a book have to be fixed to be a book? Doesn’t the discipline of finishing — of committing to a version, sending it out, defending it for life — separate serious work from endless tinkering? The objection has weight, and I take it seriously. So let me answer it.

The book object ships. On January 27, 2027, two finished, fixed, professionally produced books leave the press. They have ISBNs. They will have audiobook editions. They will be reviewed and criticized and shelved. Anyone who reads the January 2027 edition of Built to Extract will be reading a specific, signed, defensible argument that I will stand behind for life. The fixedness of the book object is preserved. What is new is the relationship between that fixed object and the thought that produced it. The thought keeps moving. The next edition, when it comes, will be a different fixed object — also signed, also defensible — that reflects what the moving thought has learned between editions. The book is fixed. The argument is alive. Both are true at once.

This is the same move I keep asking manufacturers to make with their workers. The line is fixed. The shift starts at six and ends at two. The standard work is the standard work. What is alive is the worker’s intelligence inside the system — the suggestions, the catches, the small improvements that compound into a workforce that gets better every year while doing the same job. Capital and labor as a twin helix, each ascending with the other. The book and the platform as a twin helix, each refining the other. The publishing model rhymes with the manufacturing model because both are formed by the same conviction: the artifact is not the thinking. The artifact is the most visible product of the thinking. Treat the thinking as the asset, and the artifact compounds. Treat the artifact as the asset, and the thinking dies the moment the artifact ships.

That is why the books we are publishing through Capability Capital Press will not die at the moment they ship. The mechanism is in place. The discipline is running. The thought is alive. The Monday morning answer is exactly what the book is built to deliver, because the book is being kept honest by every Monday morning until it goes to press, and by every Monday morning after.

The fossils still have wisdom in them. I read them and I love them. But I am not going to write one.

———

Venki Padmanabhan is the founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press. He writes at The Long Game.

Paul Osterman has written the diagnosis we needed. In a Bloomberg Businessweek essay drawn from his forthcoming Harvard University Press book, he documents what he calls the arm’s-length workforce: a concierge employed by a contractor rather than the building, a travel nurse employed by a staffing firm rather than the hospital, a document-discovery attorney employed by no one in particular. More than one-third of US workers, by his count, now have an employer with little to no stake in their future. He surveyed 6,141 adults to land the number. He calls the phenomenon disposable work.

The data is right. The consequences he names — economic mobility stalling, political polarization deepening, public health eroding — are right. The word is wrong.

Disposable describes what the employer does. It puts the verb on the firm and the noun on the worker, and it leaves us arguing about whether the firm should have shed less, or shed more carefully, or shed under different rules. That’s the labor-market debate we’ve been having for forty years and it has not moved.

The word he almost used is unformed. A worker the employer has no stake in is a worker the employer is not forming. Disposability is the symptom; absence of formation is the disease. Once you name it that way, the question stops being how to regulate the shedding and starts being who is supposed to do the forming — and what we owe a person in the first thirty years of a working life.

That’s the argument.

Mobility Without Rungs

Osterman’s first claimed consequence is economic mobility. He is right that the ladders are gone. He is one frame short of explaining why.

Ladders are not the mechanism. Rungs are. Rungs are built by someone — a supervisor who notices, a journeyman who corrects, a firm that keeps a record of what a worker now knows that she did not know last year. The arm’s-length employer does none of this. The worker moves between assignments, and nothing she learns travels with her except in her own unverified memory. That is not a mobility problem. It is a formation vacuum dressed up as a mobility problem. The Capability Account — an individual-owned lifetime ledger of verified capability, with mandatory employer deposits at years one, three, and five — is the missing infrastructure.

The Body’s Slow Verdict

A person in the first thirty years of working life is owed formation by four institutions: the Home that raises her, the School that tracks her, the College or Vocation that certifies her, and the First Employer who deploys her capability and adds to it. These are the Four Formations. Three still exist in some form. The fourth has been quietly dismantled — and Osterman has given us the number.

Consider Lukas. Two parents at home. The German school track. A full forty-two-month Ausbildung, Geselle at nineteen, Meister at twenty-six. Siemens as his first employer for five years. By thirty he had roughly twenty-five thousand hours of formed work behind him, recorded by institutions with standing to record it. Lukas is not exceptional. Lukas is what a working life looks like when all four gardeners stay at their posts.

The arm’s-length workforce removes the fourth gardener at scale. Disposability is a public health crisis because being unformed is a public health condition. The deaths-of-despair literature has been circling this for a decade without quite naming it.

Polarization Is Downstream

Contemporary political anger is not primarily a demand for higher pay. It is a demand to be recognized — treated as someone whose work matters to an institution larger than the self. Formation is exactly that recognition, delivered over time by institutions with standing to deliver it. When the first employer becomes a staffing intermediary with no stake in the worker’s thirty-year arc, the recognition channel closes. The worker is paid. She is not seen. Polarization is what an unformed worker does with the absence of being seen.It is not the wage. It is the witness.

Formation Is Not Charity. It Is the Better Investment.

Here is where Osterman’s frame and mine part company most sharply. The labor-market tradition treats employer investment in workers as a cost to be minimized, a benefit to be regulated, or in its most generous register a moral obligation to be encouraged. All three concede the central point to the disposability frame: that formation is something the firm does for the worker, against its own interest.

That is wrong. Formation is the better investment. It is the comparison case to automation, and on the operator math it wins — not by a little, and not in one year, but by orders of magnitude over time, because formation compounds.

The mechanism is straightforward. A formed worker in year one is perhaps twenty or thirty percent more productive than an unformed one. On its own, that does not beat automation, which delivers in the same year. But capability does not depreciate at the rate of a machine; it compounds at the rate of deployment. The formed welder in year three carries year one’s capability plus what years two and three added. Verified — recorded, witnessed, attested — it travels with her and across the workforce. One formed welder teaches the next.

Siemens Amberg is the textbook case. The Electronics Works in Bavaria has operated with roughly 1,100 employees since the late 1980s. Over the same period, production has grown eightfold, quality has reached 99.9999%, and the plant now produces 1,200 product variants with 200 new products introduced each year — at one product per second, 350 changeovers per day. Same headcount. Different category of return. Professor Karl-Heinz Büttner, who ran the plant, put it in one sentence: “Machines don’t come up with ideas for improving the system itself.”

Lincoln Electric is the older American case and the more pointed one for the disposability debate. The Cleveland arc-welder has not laid off a qualifying employee in seventy-seven years. Workers share in profits, average roughly $73,000 a year, and stay for decades — which means today’s senior welders carry process knowledge the firm has paid to develop and paid to retain. Lincoln’s productivity per worker is famously several times the industry mean. Famously profitable across cycles in which competitors are not. Famously, in the 1990s, it forced General Electric — twenty times its size — out of the welding business entirely. The firm did not buy that position. It compounded into it, one formed welder at a time, for eighty years.

Royal Enfield is the recent proof in a sector that had given up. Same workforce, same factories, different belief system: production grew from roughly fifty thousand units to one hundred and thirteen thousand, and profit grew twentyfold, without disposing of the workers the previous management had treated as the problem. The workforce was not the problem. The absence of formation was.

What the gardener does — what makes formation compound — is named in three motions. Provide the mud: psychological safety, the conditions to make a mistake without being shed. Provide the water: the steady training, tooling, and mentorship that keep capability growing. Provide the sun: the demanding work itself, the deployment that converts capability into output and the worker’s own sense of having been seen. Mud, water, sun, deployed by gardeners who know what they are doing, produce bloom. Bloom compounds — and the machine you bought in year one is depreciating while the worker you formed in year one is still teaching the worker you formed in year three.

The arm’s-length employer does none of this. Not because it is cruel, but because it has decided — at the level of capital allocation — that the formed worker is not worth the investment. It has performed the comparison on year-one math and stopped there. Osterman’s one-third is the population on which the misallocation has been performed.

What the Firm Is For

Osterman will likely close, when the book lands, with the canonical labor-market remedies: joint-employer liability, portable benefits, sectoral bargaining, stricter classification of independent contractors. These are not wrong. They are one floor too high.

The floor below is the question of what the firm is for. A firm that exists only to extract labor from a workforce it did not form is a firm that has outsourced its civilizational function and kept only its financial one — and it has done so on a capital-allocation thesis that does not survive contact with Amberg, Cleveland, or Chennai.

The right noun is not disposable. It is unformed. The right question is not how to regulate the arm’s-length employer. It is whether a firm that holds itself at arm’s length from the people who do its work is, in any morally serious sense, an employer at all — and whether, on the numbers alone, it is even a competent allocator of capital.

Capital is crystallized labor. Labor is capital in formation. For information to bear dividends, formation is the prerequisite. The arm’s-length workforce is the wage paid for forgetting that.

****

Venki Padmanabhan is the founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press.

Source: “Why Two Big Companies Just Cut Paid Family Leave,” Claire Cain Miller, The New York Times, May 11, 2026. https://www.nytimes.com/2026/05/11/upshot/paid-family-leave-deloitte-zoom.html

Deloitte just cut paid family leave for its support staff in half. Zoom trimmed parental leave by four to six weeks for most of its workforce. The New York Times framed this as a retreat from a “golden age of benefits,” a DEI rollback, a signal to women that maybe they should work somewhere else. All of that is true. None of it is the actual story.

The actual story is that the sanctuary was always rented. Companies extended family-friendly benefits during a tight labor market because they needed to recruit. The moment the leverage flipped, the benefits flipped. Laszlo Bock, who used to run HR at Google, says it plainly in the piece: companies are profit-maximizing machines and they will take advantage of a weak job market. The “we’re like a family” language was borrowed capital. It was never a covenant.

And that exposes something the article never names. Every party in this debate — the companies cutting, the advocates protesting, the researchers measuring optimal leave duration — is operating inside the same frame. Labor is a cost. Care is overhead. Time off is a perk the firm grants or withdraws based on market conditions.

But Time, Love, Health, and Wealth are not perks. They are the four conditions under which a human being becomes capable of contributing anything at all. They are the Mud, the Water, the Sun, and the Soil of formation. A parent who cannot recover from childbirth cannot bond. A parent who cannot bond cannot raise a child who will one day walk onto a shop floor or a hospital ward or a classroom and bring frontline intelligence to bear. Cut the leave, and you have not saved money. You have skipped a deposit into the capability account of the next generation, and the bill comes due in twenty-two years when that child shows up to work already depleted.

The Deloitte cuts are not a benefits story. They are a balance sheet error. That’s the argument.

In 2000, my wife and I were both at GM Lansing Grand River, launching the plant. I was a trim shift leader. She was a quality engineer on the same line in the same building. We had a two-and-a-half-year-old at home. And then she gave birth to twins.

Picture that year. I was working second shift. She was working first. Three children under three at home. The cars were rolling off a brand-new line where every defect found its way back to one of us. If the sanctuary had not held, there was no possibility — none — that either of us would have walked onto that floor and been the kind of deployed intelligence the launch required. The body would have shown up. The mind would have been in three places at once, none of them on the line.

It held because the conditions were there. Paid leave that was real, not symbolic. Health benefits that covered the pregnancy, the delivery of twins, and every pediatric visit that followed. We flew my parents over from India to be the additional sanctuary that tided us over — to take shifts with the children that we could not take. A company that understood, at least then, that the worker standing at the station was a whole human attached to a whole household, and that the household had to be funded if the station was going to be staffed.

We survived the two-year launch. The cars were successful. Lansing Grand River won the J.D. Power Gold Plant Quality Award. That is not a coincidence and it is not a story about my wife and I being heroic. It is a story about what becomes possible when Time, Love, Health, and Wealth are held in place long enough for two exhausted parents to bring full capability to a difficult job. The deposits made the output possible. Skip the deposits and the J.D. Power award goes to a different plant.

That is the part the Deloitte memo can’t compute.

I run a plastics plant in Wooster, Ohio. Sixty people, three shifts, machines that do not care what time of day it is. I can tell you each of their names. I can tell you which of them is bringing full capability to the floor on any given Tuesday morning, and I can tell you why when they aren’t.

It is almost never about training. The training is done. The certifications are current. The standard work is clear.

What varies is whether the human standing at the machine got to sleep, whether someone at home is sick, whether the daycare called at 2 a.m., whether the rent went up again, whether a parent is dying in another state. Time, Love, Health, Wealth. When any one of those four is collapsing at home, frontline intelligence does not show up at work. It cannot. The body is there. The mind is somewhere else, doing the triage that the firm refused to underwrite.

The cost of that absence is not a line item Deloitte can find. It shows up as scrap. It shows up as a near-miss in the safety log. It shows up as a part that ships out of spec because the operator who would normally catch it was running on three hours of sleep after a newborn cried all night and there was no leave left to take. The firm saved eight weeks of paid leave and lost a customer audit. The math, properly done, was never close.

This is what labor-as-cost cannot see. It measures the deposit and never measures the return.

Now extend it to the Deloitte support staff. The Times piece notes, almost in passing, that the cuts target jobs that pay less and are predominantly done by women. Joan Williams, the Equality Action Center director quoted in the article, is right that this hits women hardest. But the deeper observation is the one she stops short of: Deloitte has decided that the formation work done by a paralegal or an administrative assistant is somehow less load-bearing than the formation work done by a consultant.

That is not a defensible claim. The paralegal who comes back from twelve weeks of leave instead of eight is not less capable; she is more capable, because her sanctuary held. The consultant whose leave was preserved is not more valuable; he is just more expensive to replace in a tight market. Deloitte is not making a judgment about capability. It is making a judgment about leverage. And it is calling the judgment a benefits decision because the alternative — admitting that the firm extracts formation value it refuses to pay for — would be unsayable in a press release.

Read the Starbucks counter-example in the same article carefully. Betsy McManus, the spokeswoman, says: “When our partners are supported, they’re better able to care for our customers.” That sentence is the labor-as-capability frame slipping out in public. Starbucks expanded retail parental leave to eighteen weeks because they expect a return on the deposit. They said the quiet part out loud. It is rare, and it is correct.

Here is what a serious accounting system would do differently.

Every American has a Social Security number — a lifetime ledger that records what you earned and what you owe. Imagine a second ledger running alongside it. Call it a Capability Account. It records what was deposited into you and what you went on to produce. Three kinds of entries: the formation work (the leave a parent took, the schooling, the care), the capability demonstrated (verified by someone other than the person who trained you, because the trainer always grades generously), and the contribution made (what you actually built, fixed, shipped, healed, taught). Three entries, one human, one lifetime.

The employer is on the hook for deposits at year one, year three, and year five. Not as charity. As accounting. The firm is the proximate beneficiary of the formation work that produced the worker standing at the machine, and the firm is therefore the proximate party responsible for recording the deposit. You used the asset. You book the cost.

Under that architecture, the Deloitte cut stops being a benefits decision and becomes something the firm cannot quietly do. It is a refusal to make a required deposit, and it shows up as a missing entry on the ledger that follows the worker for the rest of her career. The next employer reads the ledger. The market reads the ledger. Deloitte’s reputation reads the ledger. The savings the firm thought it booked are immediately repriced, in public, by every party that touches the worker afterward.

Time, Love, Health, Wealth are not benefits. They are the soil. The sanctuary was rented because we let it be rented — because we accepted, for forty years, an accounting system that could see the cost of leave but not the cost of its absence. Deloitte and Zoom are not the villains of this piece. They are the symptom. The villain is the ledger that lets them book the savings without booking the depletion.

Frontline intelligence — the kind I watch walk onto my shop floor every morning, the kind every economy ultimately runs on — cannot be deployed by humans whose sanctuary is collapsing. The deposit is the precondition. Skip it and you do not get a cheaper worker. You get a depleted one, and eventually you get no worker at all.

The bill always comes due. The only question is whether the firm that benefited is the one that pays it.

***

Dr. Venki Padmanabhan is a 36-year manufacturing leader (GM, Chrysler, Royal Enfield, Ather, ADS) and founder of the Capability Capital Institute. His books Built to Extract and Already Paid For are forthcoming from Capability Capital Press in January 2027.

Subscribe to The Long Game on Substack: https://thelonggameforall.substack.com

It is Memorial Day, and the country pauses to read names on walls.

This morning the Vatican released Pope Leo XIV’s first encyclical, Magnifica Humanitas — “Magnificent Humanity” — on safeguarding the human person in the time of artificial intelligence. Two hundred and forty-five paragraphs. I have tried to read it carefully today.

The encyclical opens with Nehemiah. The exiles returning to a Jerusalem in ruins, the walls collapsed, the gates burned. Nehemiah does not impose a solution. He convenes the families. He assigns each of them a section of the wall — men, women, priests, artisans, heads of households, young people. Each section is small. Each family knows its stones. The city is rebuilt because everyone is given a section.

I have spent thirty-six years on plant floors. The floor is a section of that wall. That’s the argument.

* * *

His Holiness has named, in the magisterial citation chain I do not possess, much of what I have been trying to name from the floor.

Paragraph sixty-seven extends the Church’s ancient principle of the universal destination of goods — that creation is given to all and must not be hoarded by a few — to patents, algorithms, digital platforms, technological infrastructure, and data. When these remain concentrated in the hands of a few, a new imbalance is created. This is the same complaint Built to Extract makes. He gives it five centuries of doctrinal weight in three sentences.

Paragraph one hundred and nine recasts the five principles of Catholic Social Doctrine as criteria for AI governance: name the new monopolies; ensure universal access to technologies and the education to use them; protect communities’ ability to choose, not merely to be consulted after the fact; recognize the hidden workers who sustain the algorithmic stack; question the global distribution of power that decides who trains the models and who is subjected to them. This is, sentence for sentence, what the Twin Helix has been trying to diagram.

Paragraph one hundred and seventy-three names the hidden chain — the data labelers, the content moderators, the children in cobalt pits — as the contemporary face of trafficking. His words: their bodies are scarred, injured, and worn down so that computational flow may continue uninterrupted.

That is the labor-side ledger. Time, love, health, wealth — depleted so the capital helix could compound. The encyclical does not use those four words. It does not need to. The depletion is named.

* * *

Memorial Day is the holiday that asks who paid the price for what we have. The wall of names is a wall of formations interrupted by war. The floor has its own such wall, interrupted by something quieter. On this day I want to read four names from the section I know.

Debra Fogle was a paint inspector at the Rehau plant in Alabama. She caught defects no fixture saw. She trained the next inspector with the patience of someone who knew the line could not lie about what came off it.

Wally Vinton ran wiring harness build at the Lansing Grand River Trim Shop. The harness is the nervous system of a car. Wally taught his crew what a clean route looked like — not as a standard, but as a discipline of attention.

Ramon Hernandez was, by any honest accounting, the best team leader I ever watched. The shift settled around him without his raising his voice. People who had been written off elsewhere came alive on his crew.

Dennis Boutwell was the first supervisor I hired when Lansing Grand River opened. He is now Director of Manufacturing Excellence at Advanced Drainage Systems.

These four are not anecdotes. They are witnesses. Each carried the four depletions on their shoulders so the capital side of the helix could compound: time given to shift work that was never returned, learning paid for in their own evenings, health discounted in the actuarial tables of cumulative exposure, wages that did not keep pace with the wealth their formation underwrote. They were not enslaved. They were Americans with choices. But the encyclical’s paragraph one hundred and seventy-three reaches further than slavery in its narrow sense — it reaches toward every chain where formation is extracted without acknowledgment.

Each of them, a section of wall. Most of those sections went uncounted.

The encyclical’s deepest Memorial Day word, for me, is in paragraph one hundred and fifty-four: a paradox of material progress and anthropological regression. The thing built rises. The person who built it is not always lifted with it.

* * *

His Holiness, faithful to his office, declines to propose technical instruments. Paragraph twenty-four is explicit: Social Doctrine is not a repertoire of technical solutions. He sets the criteria. He leaves the rooms to be built by those who know the trades.

The Capability Account is one such instrument. A portable lifetime ledger that records the formation each person accumulates — hours, certifications, mentor witness, demonstrated competencies — and travels with them. Employers deposit at years one, three, and five. The ledger depreciates on a half-life so it stays honest. It is accredited, in the architecture we are building at the Capability Capital Institute, by a body that sits outside any single firm. Dennis’s arc — from first supervision in 2001 to running excellence across a public company in 2026 — is what the Account would have made portable for the thousand people on his crews who did not have his luck of being noticed.

The Five Gardeners — parents, teachers, vocational masters, first employers, and the institutional architecture that makes formation portable — is one distribution of responsibility. Subsidiarity from paragraph seventy-one, applied to formation: the decisions closest to the person are made closest to the person. The exemplars are not abstractions. I learned what a teacher is from De Nobili School in Dhanbad, named for the Jesuit who walked into Madurai in 1606 and learned Tamil and Sanskrit before he presumed to teach. Ignatius of Loyola, whose Spiritual Exercises shaped that school and a thousand like it, gave the West its first systematic discipline of formation — a four-week ledger of attention, kept by the formed against themselves. The Five Gardeners are not metaphor. They have exemplars. The exemplars have names.

None of this is the Catholic answer to the encyclical. The Church does not need a Bloom doctrine. Bloom is one practitioner’s offering, made in the synodal mode His Holiness himself invited in paragraph twenty-seven: shared discernment. The wall, in stones.

* * *

Memorial Day is the holiday of the wall of names. Each name is a formation interrupted — what could not be completed because a young person walked toward the end of their own becoming before it could be finished.

The walls we are building now — the data center wall, the algorithm wall, the model-weight wall — risk becoming Babel’s wall in His Holiness’s reading: one language, one technology, one direction, the project that sacrifices human dignity for efficiency and aspires to reach heaven without God’s blessing. The Memorial Day question is whether the names on the floor — Debra, Wally, Ramon, Dennis, and every name like theirs — will appear in those walls’ accounting, or whether their formation will once again be the silent input that does not show up on any ledger.

The encyclical closes with the Magnificat. Mary’s song. God lifts up the lowly. The instruments Bloom proposes — the Account, the Gardeners and their exemplars — are one practitioner’s attempt to translate that lift into the language of plants and payrolls and ledgers. Not because the doctrine needs translating, but because the floor needs the doctrine to take a form it can carry.

I have a section of the wall. I know my stones.

That’s the argument.

— Dr. Venki Padmanabhan

Source: Goolsbee, A. & Syverson, C. (2023). “The Strange and Awful Path of Productivity in the U.S. Construction Sector.” NBER Working Paper, University of Chicago Booth School of Business.

Every industry I’ve examined in this series shows intelligence suppression producing measurable damage. Manufacturing: NUMMI’s worst-to-best reversal. Healthcare: 98,000 preventable deaths a year. Hospitality: 70-80 percent annual turnover. Retail: $262 billion in lost sales from poor customer experience.

But no industry demonstrates the cumulative cost of suppression more dramatically than construction. Because in construction, the damage isn’t just visible in the current year’s numbers. It’s visible across half a century of declining productivity — a trend so anomalous that economists have struggled to explain it.

I think the explanation is sitting on every jobsite in America, holding a tool, waiting to be asked what they know.

The Strange and Awful Path

In 2023, economists Austan Goolsbee and Chad Syverson of the University of Chicago’s Booth School of Business published a paper with a title that tells you everything you need to know: “The Strange and Awful Path of Productivity in the U.S. Construction Sector.”

Their findings:

A construction worker in 2020 produced less than a construction worker in 1970. The value added per worker declined approximately 40 percent over 50 years.

This is not a measurement problem. The researchers tested for deflator bias and other statistical artifacts. They examined physical measures of productivity — housing units built per worker — and found similar stagnation or decline. The trend is real.

Every other major sector improved. Manufacturing productivity more than tripled. Agriculture soared. Even government services showed gains. Construction went backward. In the auto industry, output per worker went from roughly 5 cars per employee per year in 1939 to approximately 25 by 2020. Construction moved in the opposite direction.

Capital investment wasn’t the problem. The construction sector invested in capital at rates comparable to the overall economy. The money was being spent. It just wasn’t producing results.

The Federal Reserve Bank of Richmond, in a 2025 analysis, estimated that construction productivity could be 60 percent higher if construction firms operated at the scale of manufacturing firms. McKinsey estimated that the sector’s productivity could increase by up to 60 percent through widespread innovation adoption — and explicitly noted that “technology alone will not address poor productivity” and that “a fundamental culture change” is needed.

A $13 trillion global industry, going backward for half a century. The explanation is not capital, not technology, not measurement error. The explanation is suppression — an operating model that structurally prevents the people building the project from improving the design of the project. That’s the argument.

Why Construction Is Different

The suppression mechanism in construction operates through a distinctive architecture that makes it, in some ways, even more entrenched than Taylor’s factory model.

Adversarial contracting. The dominant procurement model — design-bid-build — separates the people who design the project from the people who build it, then pits the builders against each other on price. The lowest bidder wins. This creates an adversarial relationship from day one, where the contractor’s primary incentive is to minimize cost (including labor investment) and maximize change orders when things go wrong.

In this model, the trade worker — the carpenter, electrician, ironworker, or heavy equipment operator — has no formal connection to the design process. By the time the worker sees the drawings, the decisions have been made. If the design creates a constructability problem that the experienced tradesperson would have spotted in five minutes, the system has no mechanism for that intelligence to reach the designer. The worker adapts, works around the problem, and the cost shows up as a schedule delay or a change order — both of which are attributed to execution, not design.

Fragmentation. Construction is one of the most fragmented major industries. A single project may involve dozens of subcontractors, each managing their own workforce, each focused on their scope, with minimal coordination across trades. The institutional knowledge that accumulates within one project — the sequencing insights, the coordination solutions, the material behavior observations — walks off the site when the project ends and doesn’t transfer to the next one.

The knowledge drain. One in five construction workers is over 55. The industry has roughly half a million field vacancies in the U.S. alone. An entire generation of tacit knowledge — developed over careers of building things — is retiring. And because the system never created mechanisms to capture that knowledge, it’s disappearing.

The experienced ironworker who knows how a specific steel connection behaves in cold weather, the electrician who can read a panel layout and predict where the interference will occur, the superintendent who has seen a hundred foundations poured and knows what the soil conditions mean for the schedule — none of this knowledge exists in any database, any document, any system. It lives in the workers. And the workers are leaving.

The Constructability Gap

There is a specific, measurable failure mode that captures the suppression pattern in construction perfectly: rework.

Rework — doing work over because it was done wrong the first time, or because the design didn’t account for field conditions, or because trades interfered with each other due to poor coordination — consumes an enormous percentage of construction labor hours. Estimates vary, but studies consistently place direct rework costs at 5 to 15 percent of total project cost, with some analyses of indirect costs pushing the figure much higher.

The majority of rework is preventable. It stems from design errors that a trade worker would have caught, coordination failures that a foreman could have anticipated, and material or method choices that someone with field experience would have questioned.

But the operating model — adversarial contracting, hierarchical decision-making, separation of design from execution — prevents that knowledge from entering the decision stream until after the error has been committed and the money has been spent.

This is the suggestion gap writ large. Toyota gets a million suggestions per year because the system asks for them. Construction gets rework because the system is structurally designed to prevent the people building the project from improving the design of the project.

The Deployment Proof: Integrated Project Delivery

The evidence that changing the operating model transforms outcomes in construction comes from Integrated Project Delivery (IPD) and related collaborative approaches.

In IPD, the traditional adversarial model is replaced with:

Early involvement. Contractors and key trade partners are brought into the design process from the beginning — when their field knowledge can actually influence the design, rather than after the drawings are complete.

Shared risk, shared reward. All parties share in cost savings and cost overruns, aligning incentives across the entire project team. The contractor doesn’t benefit from change orders. The designer doesn’t benefit from ignoring constructability. Everyone benefits from getting it right.

Frontline intelligence integration. Foremen and trade workers contribute to planning, sequencing, and problem-solving during design, not just during execution. Their tacit knowledge of how work actually gets done enters the decision stream before the concrete is poured.

Projects using IPD and collaborative delivery consistently outperform on cost, schedule, and quality. The Last Planner System — a lean construction method that puts planning authority in the hands of the foremen and trade workers who understand the actual work sequences — shows similar results.

The mechanism is identical to NUMMI, to Magnet hospitals, to Costco, to the Ritz-Carlton. When the system invites intelligence from the people closest to the work, outcomes improve dramatically. When it doesn’t, the industry goes backward for fifty years.

The Generational Emergency

There is an urgency to construction that other industries don’t share.

Manufacturing can (theoretically) rebuild frontline capability over time. Healthcare has nursing schools continuously producing graduates. Retail and hospitality can rehire and retrain relatively quickly.

Construction is losing an irreplaceable generation. The workers retiring now carry knowledge that was accumulated over 30- and 40-year careers — knowledge of materials, methods, conditions, and failure modes that exists nowhere else. The industry didn’t create apprenticeship systems robust enough to transfer it. It didn’t create documentation systems sophisticated enough to capture it. It certainly didn’t create operating models that valued it enough to systematically preserve it.

When the last ironworker who knows how to read that specific type of connection retires, that knowledge is gone. When the superintendent who has managed 50 complex pours in difficult soil conditions leaves, that judgment is gone. No amount of AI training data will recover what was never recorded.

This is the suppression tax compounded across a career, multiplied by a generation, applied to a $13 trillion global industry. The intelligence was there for decades. Nobody asked for it. Now it’s walking out the gate.

The AI Bridge

Construction AI is arriving — BIM, scheduling optimization, safety monitoring, site analytics, prefabrication robotics. The dominant narrative is that technology will solve the labor shortage.

The CCI argument is different. The labor shortage isn’t primarily a body shortage. It’s a knowledge shortage. And the solution isn’t to replace the retiring workers with robots. It’s to deploy AI as a mechanism for capturing and amplifying the tacit knowledge those workers carry before they leave — and then using that AI-augmented intelligence to train the next generation.

The picture is concrete. An apprentice learning from a tool that was trained not on textbook procedures but on the accumulated judgment of the master craftspeople who came before. The experienced electrician flagging a constructability issue from the field, with context and photos, routed to the design team before the wall is closed. AI as a frontline amplifier, in the one industry that has been going backward for half a century.

It requires the industry to do what the adversarial contracting model structurally prevents: ask the people holding the tools what they know, capture it, value it, and build on it.

The jobsite has been waiting fifty years for someone to ask.

———

Next week: “The Walden Pond Testimony” — I stop presenting other people’s data and tell you what I saw. Night shifts at a GM assembly plant. What I witnessed, why it changed everything I thought I knew, and what it means for the argument I’ve been building.

———

Dr. Venki Padmanabhan is a plant manager with 36 years of global manufacturing leadership experience, including executive roles at GM, Chrysler, Mercedes-Benz, Royal Enfield, and Ather Energy. He holds a PhD in Industrial Engineering from the University of Pittsburgh. He is the author of the forthcoming Already Paid For: Why Unlocking Frontline Intelligence Beats Automating Workers Away and co-founder of the Capability Capital Institute. He writes The Long Game at thelonggameforall.substack.com.

Source: “Flying the unfriendly skies: A business ethicist says goodbye to Spirit Airlines,” by John Paul Rollert, Fast Company, May 12, 2026.

By noon on May 2, 2026, the Marine Air Terminal at LaGuardia was almost silent.

Cibo Express closed half a day early. There were no customers to serve. The last TSA officer was sent home. Spirit Airlines, which had flown out of that Art Deco terminal — the one built in 1940 for Pan Am’s Clippers — was over. By dawn that morning, the website read: To our Guests: all flights have been cancelled, and customer service is no longer available.

Seventeen thousand people lost their jobs that weekend. Five thousand flight attendants. Pilots. Mechanics. Dispatchers. Ground crews. The union of flight attendants asked the federal government for a $600 weekly supplement, because the airline they had worked for could not be relied upon to pay them what it already owed.

At the bankruptcy hearing on Tuesday, the company’s attorney offered the official cause of death. A fuel “megaspike,” he said, driven by the U.S.-Israel war with Iran, projected to add hundreds of millions of dollars to the year’s liquidity needs. The HR letter from Vice President Suzanne Solon explained that the company could not give workers more notice because notice itself would have killed the rescue capital they were still trying to raise.

This is the airline’s story of its own death. It is the story you will read in most places. It is not the truth.

A healthy airline survives fuel spikes. They happen every cycle. American, United, Southwest, and Delta are flying through the same Iran war, the same fuel chart, the same year. None of them hedge fuel either. Spirit hadn’t hedged since 2015 — and neither has anyone else at scale.

The difference isn’t hedging. The difference is what the other airlines had that Spirit did not: reserves of every other kind. Balance sheet cushion. Creditor goodwill. Passenger loyalty that could be monetized in a crunch. Union flexibility that could buy time. Operational slack that could absorb shock. Spirit had stripped all of them out, and called the stripping discipline.

Spirit’s own restructuring plan, filed weeks before the shutdown, assumed jet fuel at $2.24 a gallon for 2026. By the end of April, the price was $4.51. The forensic firm Santiago & Company calls this an “optionality failure” — efficiency without reserves, a structure with no slack. Every dollar of fuel cost above plan translated to $109M in cash pressure. The company tried to raise $500M in rescue capital. No one came.

No one came because no one was left to come.

Not the passengers, who had been treated as fees-to-be-extracted for fifteen years. Not the employees, whose interactions had been scripted by the Disney Institute in 2017 because Spirit understood — correctly — that the frontline mattered, but understood — incorrectly — that the answer was to train the smile rather than deploy the judgment. Not the creditors, who had watched the model and knew there was nothing left in the tank to lend against. Not the regulators, not the unions, not the public.

This is what extraction does over time. It strips the relational reserves a business needs to survive a normal industry event.

Ben Baldanza, who ran Spirit from 2005 to 2016 and died of ALS in 2024, called what he built a “bus with wings.” He was proud of the phrase. The Wright Brothers Memorial Trophy committee, awarding him in 2024, said the model created “an avenue for middle-class and working-class Americans to fly to leisure destinations.” Both descriptions are true. Neither is complete.

What Baldanza built — and what the University of Chicago business ethicist John Paul Rollert has been documenting for eleven years — was a system designed to identify every dignity that could be unbundled, priced, and resold. Snacks. Seat choice. Carry-on bags. Boarding order. Water. The bare fare was not, as Baldanza insisted in his 2015 Booth debate with Rollert, “about saving our customers money.” It was about discovering how little a customer could be given and still be called a customer.

The same logic ran inside the company. The flight attendants who serviced the planes. The mechanics who maintained them. The gate agents who absorbed the rage of passengers who had just discovered their carry-on cost $89. Every interaction designed for minimum cost. Every worker rated on transaction volume rather than judgment. Every complaint logged as data for the next fee, not as a signal that the model was eating itself. This is the BTE pattern in its purest form: the system suppresses intelligence and extracts value. The intelligence was everywhere — in the gate agent who knew which flights were about to go wrong, in the mechanic who knew which planes needed attention, in the flight attendant who knew which routes were losing the customers worth keeping. Spirit had all of it. The model required that none of it be used.

Sara Nelson, president of the Association of Flight Attendants-CWA, wrote to Transportation Secretary Sean Duffy and acting Labor Secretary Keith Sonderling the day after the shutdown. She asked for two things. That the flight attendants be paid the wages and per diems they had already earned. And that they receive a $600 weekly federal supplement to state unemployment.

Read that letter twice.

The union of a recently-deceased airline had to ask the federal government to backstop the wages of workers whose employer had been profitable for years. The public was always the unwritten subsidy. The fare was loaded — onto employees who absorbed the abuse, onto passengers who paid the fees, and now onto taxpayers who would cover what the airline could not.

Jason Ambrosi, president of the Air Line Pilots Association, said it directly: “The pain of this decision will not be felt in boardrooms. It will be felt by pilots, flight attendants, mechanics, dispatchers, and ground crews, and by the families and communities that depend on them.” This is the BTE thesis in fourteen words, written by a pilot’s union president in the airline’s obituary. You do not need to argue it. You only need to read it.

And here is where the obituary turns.

Spirit Airlines is dead. The yellow planes are gone. The Marine Air Terminal is quiet. But the model — the bare fare, the unbundled dignity, the worker-as-cost, the passenger-as-fee-stream — is now the industry standard. Basic Economy on Delta. Basic Economy on United. Basic Economy on American. Every major U.S. carrier now charges for what Spirit pioneered charging for. The race to the bottom, as Rollert calls it, has been won. The bottom is now the floor.

Spirit didn’t die. Baldanza won.

And this is what makes the obituary necessary. Because the public conversation will treat this as the end of one badly-run airline. A fuel spike, a bankruptcy, a sad day for 17,000 workers. It is not the end of anything. It is the moment at which the extraction model became invisible — absorbed into the operating logic of every carrier in the country, no longer a Spirit specialty but an industry baseline.

What was the alternative? The alternative was already paid for. The intelligence was already on the planes. The gate agent who could have saved the customer. The mechanic who could have flagged the route. The flight attendant who could have built the loyalty that, in the end, no rescue investor could find. Spirit had all of it. The model required that none of it be used.

A healthy airline could have spent the last decade building the reserves — financial, relational, human — that survive a normal fuel cycle. Spirit spent the same decade stripping them. The body was strong. The route network was real. The brand recognition was real. The 50,000 passengers carried on the final Friday were real. But there were no walls. A normal wind blew it down.

———

Venki Padmanabhan is the founder of the Capability Capital Institute and the author of Built to Extract and Already Paid For, forthcoming from Capability Capital Press.

A maintenance supervisor retired four years ago. Thirty years at a plant in Alabama. He had started as an operator on the molding line, worked his way up, and over three decades had installed every piece of major equipment in the building. He knew the sound a gearbox made forty-eight hours before it seized. He knew which electrical panels had been wired by which contractor and where each one had buried its shortcuts. Thirty years of institutional knowledge in his hands, none of it in a manual.

When he retired, the HR file recorded a benefits payout, a final paycheck, and a thank-you card signed by the shift. The general ledger recorded nothing else. No write-down. No impairment. No depreciation schedule closing out. As far as the accounting system was concerned, nothing of value had left the building.

Two quarters later, the machines began to crash in ways no one on the floor could explain. Throughput dropped. The new maintenance lead blamed the supplier, the heat, the schedule. He did not blame the ledger, because the ledger never told him there was anything to blame. The asset that walked out the door was never on the books in the first place. An asset stranded at birth — recorded nowhere, consumed continuously, replaced almost never.

Two disciplines have brushed against this and walked away. Oana Labes posted a sharp note this week listing seven limitations of EBITDA. Her list is correct as far as it goes, and it stops where the financial accounting tradition has always stopped: just shy of naming the asset that EBITDA was designed to hide. The operational accounting tradition stopped at the same place, fifty years earlier, in a different language. Lean called it the eighth waste, named it as a placeholder, and never built the system to address it.

That’s the argument. Finance has seven EBITDA limitations and a missing eighth. Lean has seven wastes and a bolted-on eighth that no one operationalized. The convergence is not coincidence. It is the same asset — the formed capability of the frontline worker — going unrecognized in two different accounting languages. You can only manage what you measure. Both traditions abandoned the metric for the asset that mattered most. Until that asset is on the books, neither discipline can finish what it started.

What Ohno Built, and What the West Could Not Carry

Taiichi Ohno codified seven wastes in the Toyota Production System: transportation, inventory, motion, waiting, overproduction, overprocessing, defects. TIMWOOD. Each waste came paired with a method to eliminate it — kanban for overproduction, jidoka for defects, one-piece flow for waiting. They named what happened to the work, not what happened to the worker’s capability — because at Toyota, capability was not a separate category. The plant was, in Ohno’s framing, a Thinking People System. Workers stopped the line. Workers proposed kaizen. Respect for the human was not a waste category. It was the substrate the whole system rested on.

Then TPS crossed the Pacific. American practitioners in the 1990s — Jeffrey Liker, Norman Bodek, the academic translators of The Toyota Way — added an eighth: non-utilized talent. Unused human potential. The waste of skills. It went on the posters. It went into DOWNTIME, the American acronym that displaced TIMWOOD.

The eighth was added because something visible at Toyota was not transferring. The Western adopters could see the seven process wastes did not account for the human substrate. So they named it. But they could not carry what they were naming. Ohno did not need an eighth waste — the seven, applied through autonomation and continuous improvement, were already the operational expression of respect for the worker. Strip the substrate out, paste the wastes onto an American assembly line that still treats labor as a substitutable input, and the seven can be eliminated while the eighth grows. The poster goes up. The capability runs down. No one notices because no one has the instrumentation to measure it.

Substrates do not bolt on.

What Malone Built, and What Finance Could Not See

John Malone took over Tele-Communications Inc. in 1973. The cable industry he inherited was capital-intensive and structurally unprofitable on a GAAP basis. The lenders he needed would not underwrite against losses. So he invented a metric. Earnings before interest, taxes, depreciation, and amortization. He argued — successfully — that depreciation was an accounting fiction, that the cash flow was real, that the business should be valued on what it generated before the bookkeeping took it away. He targeted five times debt to EBITDA. The leveraged buyout era picked up his vocabulary and never put it down.

The Buffett-Munger-Klarman critique was the right one. EBITDA is not cash flow. Depreciation may be a non-cash charge in the period it appears, but it records a real economic event — the consumption of a long-duration asset that will need to be replaced. Strip it out and you are measuring the cash the business generated before paying for the assets it consumed to generate it. EBITDA makes capital-consuming businesses look cash-generative. It was designed to.

Oana’s seven limitations name what the financial tradition has learned to see: EBITDA is not cash flow; depreciation reflects real reinvestment; interest and taxes don’t disappear; EBITDA ignores working capital; “adjusted” usually means “inflated”; EBITDA inflates debt capacity; it was built for bankers, not operators. Each has a partial countermeasure. The financial tradition has built tools for the first seven. It has not built one for the eighth.

The eighth, in financial language, is the same asset lean could not transfer. A capable frontline worker is a long-duration asset by every test we apply elsewhere — substantial cost to form, long economic life, real depreciation as skills erode and technology shifts, continuous consumption that must be replaced through ongoing formation or it runs out. The press is on the balance sheet. The worker is not.

The consequence is Malone’s trick, applied to a different asset. The firm consumes the accumulated capability of its workforce — drawing down the stock formed by schools, prior employers, families, and the worker’s own years of practice — and reports the output as EBITDA. The acquirer bids on the metric. The capability runs out two owners later. The plant closes. No one finds the cost on any P&L because it was never booked anywhere.

The Operator Knows Before the Number Does

I tried to bring the supervisor back. The math was straightforward. Outside maintenance contractors were costing the plant $180,000 a year for emergency calls. Bringing him on full-time at the rate he was asking would cost $104,000. Conservative ROI estimate, including throughput recovery and scrap reduction: four to one. I built the case and walked it up.

The answer came back from regional management. The existing maintenance supervisor on staff was paid $40 an hour. HR policy did not allow a technician to be paid more than a supervisor. Maximum offer: $42 an hour. The retired supervisor had asked for $50.

Eight dollars an hour. About seventeen thousand a year, against four hundred thousand of estimated annual value creation. The 4:1 ROI proposal was declined because the firm had no asset on the books to defend, no carrying value to compare the rate against, and no language for what was being lost. The plant manager — me — knew. The CFO did not. The regional manager did not. The lean consultant who had hung the eight-waste poster in the breakroom three years earlier did not. The poster named the waste. It did not give the regional manager an instrument to measure what he was declining.

EBITDA was built for the side of the table that does not have to live with the consequences of asset consumption. The operator who has to run the line in year seven is reading the wrong number. The eighth waste poster, hung in the same breakroom where the retirement card was signed, named what was leaving without offering a way to hold onto it.

The Convergence: Solving the Wrong Problem the Right Way

Two disciplines. Two languages. Two lists of seven. Two placeholder eighths. Both traditions were built to optimize what happens to the work. Both treated the worker as the agent who performs the work, not as the asset that accumulates the capability to perform it. Neither has a category for the depreciation of human capability inside the firm, because neither has a category for its capitalization in the first place. When something is not capitalized, its consumption is invisible. When its consumption is invisible, its replacement is optional. When its replacement is optional, it does not happen.

Statisticians have a name for this failure. Richard Hamming put it cleanly: It is better to solve the right problem the wrong way than to solve the wrong problem the right way. The literature calls it the Type III error. Both lean and EBITDA are textbook Type III. Lean’s seven wastes are solved with technical precision, while the eighth — the asset itself — is left as a poster. EBITDA’s seven limitations are addressed with adjusted-EBITDA refinements and covenant carve-outs, while the eighth — the same asset — is left unmeasured. Both traditions answer their own questions correctly. Neither is asking the right question.

This is why lean’s eighth waste never got its kanban, and why EBITDA’s eighth limitation never got its line item. Both disciplines named the symptom. Neither could prescribe the treatment, because the treatment requires capitalizing the asset, and capitalizing the asset requires a third tradition both disciplines assumed someone else was handling. The third tradition is formation — and the accounting that would make formation legible. Human capital accounting. Social and cultural capital accounting. Trusted-relationship accounting. The work of measuring what EQ-driven leadership, institutional reputation, and accumulated frontline judgment actually contribute to a firm’s long-run earning power. None of it sits on a GAAP balance sheet. All of it shows up in the variance report when it leaves.

What Catches What Both Missed

Formation is not a cost. It is the act of capitalizing the asset EBITDA was designed to extract for free and lean named as a waste without knowing how to prevent.

When a firm invests in the formation of its frontline workforce — supervised practice, judgment-building events, time with capable elders — it is doing what Ohno’s plant did and what the Western export of TPS could not transfer. It is building the substrate. The Bloom system names the conditions: Mud, Water, Sun. The soil the firm prepares, the flow it sustains, the attention it gives. These are the recapitalization schedule for the asset both the lean poster and the EBITDA report leave off.

And here is where automation becomes the fork in the road. The same tool, applied two ways, produces opposite accounting consequences.

The Built-to-Extract logic uses automation to replace the stranded asset. Eliminate the wage line, book the savings as EBITDA improvement, harvest the residual capability until it runs out, sell to the next acquirer. The metric loves this. It was built for it.

The Already-Paid-For logic uses automation to augment the formed capability. Take the supervisor who has the judgment to know which gearbox is failing before the vibration sensor catches it, give him a tool that hears what he cannot, and throughput compounds. The wage line does not fall — it may rise — but output per formed worker rises faster, and the asset on which the augmentation rests is being replenished by the same Bloom conditions that formed it. EBITDA grows on a stock that is being maintained, not a stock that is being drawn down.

The distinction is epistemic before it is financial. The augmentation logic preserves what Ramamoorti calls human epistemic sovereignty — the principle that humans remain responsible for deciding what is true, what to trust, and what a number means in context. AI assists knowledge; it cannot replace the responsibility for knowing. The extraction logic violates that principle quietly: it lets the metric do the judging, accepts the output, and treats the absence of an operator who can call the gearbox in advance as a savings rather than a loss. What separates the two operations is calibrated epistemic humility — the discipline of knowing what the instrument cannot see, and escalating the decision to a human being at the boundary of the instrument’s competence. EBITDA has no such humility. Neither does the eight-waste poster. Two firms can post identical numbers, hang identical posters, and be running opposite operations. Only the operator knows. The metric was never designed to.

The Rental

There is a coda to the supervisor’s story that makes the accounting failure unmistakable.

He still comes. When the machines crash in ways the current team cannot solve, he drives over and fixes them. He charges a thousand dollars a day. The plant pays it because the alternative is a line down for three shifts. Over the course of a year the plant spends roughly a hundred and eighty thousand dollars on his emergency calls. It shows up on the variance report as “outside services,” up twenty-two percent year-over-year.

No one connects the line item to the retirement four years earlier. No one connects it to the $8-an-hour gap that lost him in the first place. The capability the firm failed to capitalize is now being rented back on the spot market at the rate the asset itself sets, and the rental shows up as an operating expense unattributed to the original failure. The firm gets the worst of both possible worlds. The consumption was invisible because the asset was never on the books. The reacquisition is visible but unattributable, because the line item where it lands has no memory of where it came from.

This is what the absence of an accounting entry costs. Not just the asset, but the firm’s ability to learn from losing it.

What Comes Next

The fix is to put the missing asset on the books. Until the formed capability of the frontline workforce has a carrying value, a depreciation schedule, and a recapitalization requirement, every firm will be reading a number that hides its most important consumption and hanging a poster that names what it cannot address. Every acquirer will bid on a stock the seller has been drawing down. Every plant manager will live with the drawdown without a way to name it. And every retired supervisor will be available, at a price, on a spot market the firm cannot afford to keep using and cannot afford to stop.

The accounting entry the economy needs does not yet exist. Building it is the work of a separate tradition — one we will take up together in the book that follows this essay.

A maintenance supervisor retired four years ago. The general ledger recorded nothing. The lean poster on the wall named the eighth waste and offered no instrument to measure it. The asset was stranded at birth and consumed in silence and is now gone — except when it returns at a thousand dollars a day to fix what no one else can, and the firm pays without ever putting the cost in the column where it belongs.

That is the argument. That is what Ohno built and the West could not carry. That is what Malone built and finance never finished. That is the right problem, asked in two languages, answered correctly by neither. That is what is still not on the books.

———

Venki Padmanabhan and Sridhar Ramamoorti are two of the eight founding principals of the Capability Capital Institute. Venki brings the operations tradition — thirty-six years of manufacturing leadership across the United States, Europe, and India, currently plant manager of the Wooster plant at Advanced Drainage Systems, and is the author of Built to Extract and Already Paid For. Sridhar brings the accounting tradition — Associate Professor of Accounting at the University of Dayton, co-author of more than sixty articles and fifteen books including The Audit Committee Handbook and The A.B.C.’s of Behavioral Forensics, and formerly a principal at Andersen Worldwide, National EY Sarbanes-Oxley Advisor, and corporate governance partner at Grant Thornton. Their first co-authored book, Accounting for Formation, is forthcoming from Capability Capital Press.