What the 2026 numbers say about the manufacturing moment, and the decision nobody is naming.
This document has a date on it because it has a shelf life.
Every figure in here traces to a named source, and roughly half of them will be wrong within a year. Capacity auctions clear twice a year. Tariff regimes have not survived six months. A Supreme Court decision vacated the largest trade action in fifty years and a replacement took effect inside an hour.
So I am not going to revise this. When the numbers move I will write the next edition, and this one will stand as what was true in August 2026.
Second thing. I am reporting here. I am not advising. I will name the clocks and show you the arithmetic. I will not run your depreciation election, structure your financing, or file anything on your behalf. When you get to that part you want a CPA who does this every day and a lender who knows the programs. I can tell you who.
That is not modesty. It is the difference between a reporter and a liability.
What I actually do comes later in this document, and only if you want it.
If you run a manufacturing business somewhere between $20 million and $100 million in revenue, you have spent eighteen months being told what your problem is by people who have never signed the front of a paycheck.
The tariffs will save you. The tariffs will kill you. AI will replace your people. AI will transform your plant. You are already too late.
Here is what is actually true.
Almost nobody has done this yet. Census Bureau data from May 2026 puts AI use at 19.8% of American businesses under the strict definition, meaning AI running in the actual production of goods or services rather than somebody's assistant drafting an email. Among firms with 100 to 249 employees it is 32%. Among the largest firms in the country it is 37%.
That is not a warning. That is a starting gun most of the field has not heard.
And you are the reason any of it matters. Not the software companies. You. Rails only pay off if somebody runs freight on them.
Joseph Schumpeter spent a career on one observation, and almost everyone who quotes him gets the tone wrong.
Creative destruction was never a warning in his hands. It was a description of how wealth actually gets made. He watched the railroads go down across Europe and America and he noticed something the canal operators did not.
The railroad did not make the canal faster.
It made the canal irrelevant.
The canal companies were not stupid. Many of them were superbly run. They had better boats than anyone had ever built, better locks, better schedules, better crews. They spent two decades becoming extraordinarily good at a thing that was quietly ceasing to matter. And the men who understood which surface they were standing on built fortunes that lasted four generations.
That is the whole law and it has not moved in a hundred and eighty years.
You are not building a perfect propeller. You are deciding whether you are in the propeller business or the jet engine business. Those are different questions and only one of them has an answer worth having.
Kodak invented the digital sensor. They held the patent. They understood the technology better than the companies that ended them. What they could not do was decide, at the level of the business itself, which category they were operating in.
Forty-two years of pattern recognition across pharmaceuticals, defense, energy, and high-growth startups produced the same finding every time. The businesses that came through a transition were not the ones with the best technology. They were the ones whose leadership made a decision about architecture before they made a decision about tools.
The ones that skipped that step did not fail loudly. They got very good at the canal.
Everything below sits underneath these.
Both are true. Manufacturing payrolls fell for a second consecutive month in June 2026, at the fastest pace since 2009 outside the pandemic.
Capital is coming home. People are not.
Nobody is building the bridge between those two numbers, and the bridge is the entire opportunity. What follows is why the gap exists, and why it has a closing date.
Most people know about Section 179 and bonus depreciation. Those are real and they are permanent now. Equipment, machinery, fixtures, land improvements, all expensed in year one, no sunset.
Almost nobody knows about Section 168(n).
The 2025 tax act created a new class called qualified production property. If you build a facility used as an integral part of manufacturing, production, or refining, you can elect to deduct 100% of the building in the year it is placed in service. Not over thirty-nine years. In one year.
Now read the dates, because the dates are the whole thing.
Construction must begin after January 19, 2025 and before January 1, 2029. The property must be placed in service before January 1, 2031.
A manufacturing plant runs twenty-four to thirty-six months from decision to placed-in-service. Site selection, financing, and permitting run twelve to eighteen months ahead of the first shovel.
Do that arithmetic against a January 2029 construction start and you land in the same place I did.
Not 2029. 2027. And 2027 starts in five months.
There is a second edge on this and it is sharper than the first. The election is irrevocable, and a ten-year recapture period starts the day the building opens. If that floor stops being used for qualified production inside the decade, the deduction comes back as ordinary income.
So think about what you are actually signing. You get one election, on a thirty-nine-year asset, expensed in a single year, with a ten-year leash on how the floor gets used.
If that building is laid out around the way you run today, you have written your current operating model into concrete and taken a deduction for it.
Which means the architecture decision has to come before the building decision. That is not my opinion. That is a recapture provision.
One more thing, and this is the part that makes waiting expensive twice over. The Fed held at 3.50% to 3.75% on July 29, its fifth consecutive hold, with three governors dissenting because they wanted a hike. Inflation is running 4.2%, the highest in three years. Seventeen of eighteen officials judged the risks tilted upward.
A first-year deduction is worth more when capital costs more. The window is closing while the cost of money is biased against you.
This is the one nobody in your industry is talking about, and it is the one that will decide whether the other two matter.
In the PJM territory, which covers thirteen states from Illinois to New Jersey and includes most of the industrial Northeast and Mid-Atlantic, the price of generating capacity has done this:
Roughly eleven times in three years. And that $325 is not a market price. It is a ceiling negotiated by a coalition of governors. Without it, the 2027/28 auction would have cleared near $530.
Now the reason.
PJM's own 2026 long-term forecast projects 32 gigawatts of peak load growth between 2024 and 2030. Ninety-four percent of that growth is attributed to data centers.
In July 2026 the grid operator came up 6,831 megawatts short of its own reliability target. That was the second consecutive miss, and the first one was the first in the capacity market's eighteen-year history.
Paying more. Getting less.
Understand the mechanism, because it is not what most people assume. The forecast does not merely predict what gets consumed. It determines how much capacity the grid operator is required to procure. Speculative data center load goes into the forecast, capacity gets bought to serve it, and every ratepayer in the territory pays for it. Residential, commercial, industrial. Including the plant you run.
If those data centers never get built, the money is not refunded.
And the input-cost picture confirms it from the other direction. The National Association of Manufacturers found that raw material costs became the top business challenge for 83.1% of manufacturers in the second quarter of 2026, up from 57.5% the quarter before. A 25.6 point jump in ninety days.
Look at what drove it. Seventy-two percent cited rising energy input costs. Trade uncertainty ranked second and moved 1.2 points.
The thing squeezing your margins is not primarily the trade fight everyone is arguing about. It is power.
Deloitte and the Manufacturing Institute project that American manufacturing may need 3.8 million workers by 2033, and that 1.9 million of those positions could go unfilled.
There were 439,000 open manufacturing jobs in February 2026. The hardest roles to fill are the ones that take years to build: automation and controls engineers, maintenance technicians, manufacturing engineers, people who understand a process well enough to know what should be automated and what should be left alone.
Congress cannot legislate a forty-five-year-old maintenance technician into existence. Neither can a tariff.
Now here is the finding that reorganizes the entire national conversation, and I want you to sit with it.
The Reshoring Initiative surveyed more than five hundred manufacturers. Thirty percent of original equipment manufacturers said they would move production back to the United States if the workforce had higher skills and were available in abundant supply.
The same survey found that a skilled workforce is a bigger focus for those manufacturers than policy. Bigger than tariffs and taxes. Bigger than regulation. Bigger than currency.
Read that again.
The manufacturers themselves rank people above the entire political argument.
Every op-ed, every hearing, every cable segment for eighteen months has been about the variable that ranks third. The variable that ranks first cannot be hired, because the people do not exist. It cannot be legislated, because demographics do not take instruction.
It can only be architected.
I am going to handle the tariffs quickly, because they deserve less of your attention than they have taken.
Four regimes in eighteen months. The Supreme Court struck down the largest one in February 2026, six to three, after it had collected more than $160 billion. A ten percent global surcharge went up within hours and expired by operation of law five months later at its statutory limit. A replacement took effect the same minute it lapsed and was in court the week it landed.
You cannot build a cost structure on that. Nobody can. That is not a political statement, it is an observation about planning horizons.
The layer that actually reaches you never moved. Section 232 sits at fifty percent on steel and aluminum, twenty-five on copper, and an April 2026 proclamation expanded it to the entire customs value of covered articles and their derivatives, regardless of how much metal is actually in them. Industrial machinery. HVAC equipment. Automotive components. Anything with that metal inside it.
Now let me show you who won, using the U.S. Census Bureau's Quarterly Financial Report. This is the series the Commerce Department uses to calculate GDP and the Treasury uses to estimate corporate tax. It has no politics in it.
Profits before taxes, first quarter 2025 to first quarter 2026:
| All manufacturing | +34% |
| Computer and electronic products | +96% |
| All other electronic products | +119% |
| Iron, steel, and ferroalloys | +447% |
| Fabricated metal products | no characterizable change |
| Machinery | no characterizable change |
| Electrical equipment | no characterizable change |
| Food | −31% |
| Aerospace products and parts | −52% |
Manufacturing profits are up thirty-four percent. That headline has been on every business page in America.
It is being carried by two things. Companies building the hardware for the AI boom, and companies producing the metal the tariffs protect.
Iron and steel up four hundred forty-seven percent sits six rows above fabricated metal products at flat, in the same table, published by the same agency, in the same quarter.
The policy protected the mill. It taxed the shop that buys from the mill.
If you fabricate, form, machine, or assemble, you have been reading headlines about a boom you are not in. That is not a feeling. It is a federal statistic, and you have been carrying it alone because nobody put those two lines next to each other.
I have watched four of these. Three of them arrived incomplete.
The internet buildout of the late 1990s brought capital without capability. The money was there and the technology could not carry the weight yet.
The 2008 window brought capability without capital. The tools were arriving and nobody could finance anything.
What is on the table right now is capital, capability, and political will inside the same window. The capital is appropriated and sitting in statute. The capability ships today and works. The political will is bipartisan in effect if not in tone, because both parties have decided that domestic production is a national interest.
I have not seen those three converge before. Not in forty-two years.
That convergence is why this is a moment and not merely a good year. And it is why the closing dates matter so much. A window that opens this wide does not stay open, and it does not reopen on the same terms.
Every capability you build either grows durable earnings or expands the multiple those earnings are valued at. There is no third lever. Enterprise value is earnings times multiple, and right now three forces are pushing on the multiple at the same time.
The gap between zero and one has no multiple. That is the part people miss when they wait. Going from nothing to something is not an increment. It is a category change, and the businesses that make it are not twelve months ahead of the ones that do not. They are on a different surface.
And the cost of stepping onto that surface later is not the same cost. That is the Catch-Up Penalty, and it compounds in the same direction the advantage does.
Walk your floor and tell me what you see.
You do not inspect quality into a product. You build it in. You know that in your bones. Deming taught it, your industry absorbed it forty years ago, and today it is so obvious on your floor that nobody says it out loud. You do not bolt quality onto the end of a line and trust the inspector to catch what went wrong upstream. You design the process so the defect cannot happen.
You have been running that philosophy every single day for your entire career.
You are not running it with AI.
With AI you have done exactly the thing your own manufacturing discipline forbids. You bought a subscription. Somebody in the office is using a chatbot. Maybe there is a pilot. And the plan, if we are honest about it, is to inspect the results afterward and see if any of it helped.
That is not a criticism of you. It is the water everybody is swimming in. The entire market has been selling tools, and a tool is something you bolt on and inspect. Almost nobody is selling the architecture that makes the outcome structural.
But you already know the answer, because you solved this problem on the floor decades ago. Nobody has shown you that it was the same problem.
That is the whole gap. Not knowledge. Not budget. Not talent. Category recognition.
And there is a second recognition sitting right behind it, and this one is not about your company. It is about where your company sits.
Both of my grandfathers worked the anthracite mines.
My father's father worked for Kehoe-Berge, out of Pittston, a company started in 1934 by two men who wanted to put the district back to work after the big operators shut down. My mother's father worked at the Harry E Colliery in Swoyersville. He died there.
I want to be careful here, because this is not a story about hardship and I am not asking anyone to feel anything about my family. It is a story about a transaction, and the transaction is running again right now, in the same counties, on the same ground, with better lighting.
Here is how the first one worked.
The coal was under Northeastern Pennsylvania. The men were from Northeastern Pennsylvania. They went down and they brought it up and it powered the industrial rise of the United States. It ran the mills in Pittsburgh and the furnaces in Bethlehem and the locomotives that made the country a country. That coal is in the foundations of American wealth.
And the wealth was made in New York and Philadelphia.
What stayed here was holes in the ground, a river you did not swim in, black lung, and a set of towns that spent the next sixty years explaining to their children why the good jobs were somewhere else.
And the culm.
Culm is what is left after the coal is cleaned. Rock, slate, coal dust, the part with no market. The Harry E piled it in banks across a huge swath of Swoyersville, and by the time I came along in the sixties they were mountains. Not a figure of speech. Mountains, with paths worn into them.
That is where we played. Every kid in that borough played on them. It never once occurred to any of us that we were playing on the part they did not want, in a town built to hand over the part they did.
Nobody stole anything. Every contract was signed. Every load was paid for at the price agreed. The region supplied the energy for a transformation it did not own a share of, and when the transformation was complete the region was not in it.
That is the transaction. Energy out. Ownership elsewhere. Perfectly legal, and it emptied a valley.
Now, before I tell you what is being built, one more thing about those banks.
They are leaving.
Pennsylvania's Department of Environmental Protection publishes a list of active anthracite permits every year. The Harry E is on it. Permit 40-18-01, Swoyersville Borough, Luzerne County, held by Keystone Reclamation Fuel Management LLC, 19 Headquarters Plaza West, Morristown, New Jersey.
They came back for the culm.
I have watched it go out for years. Truck by truck, load by load, the mountains I climbed as a boy getting smaller, sold off by a company headquartered a hundred and twenty miles east. My grandfather's work, and then his waste, and both of them ended up somewhere else.
That is not a historical parallel I am drawing for effect. That is a permit number, current as of the most recent state filing, on the ground where my grandfather died.
The transaction never ended. It is still collecting.
Now look at what is being built.
In Lackawanna and Luzerne counties alone, developers have proposed more than one hundred data center buildings across more than a dozen campuses. A ten billion dollar campus is under construction in Salem Township. A fifteen-building campus is approved in Hazle Township. Archbald alone has five projects that would total roughly 4.7 million square feet. One proposed facility in Blakely would have required an estimated 1.5 gigawatts, which is the entire output of the gas-fired plant down the road. Pennsylvania now ranks third in the nation for proposed data centers.
Landowners in Luzerne County have been approached at a hundred seventy-five thousand dollars an acre.
The head of our regional development council called data centers the fourth industrial revolution in Northeast Pennsylvania. First it was coal, he said. Then manufacturing.
He is right about the sequence. That is what worries me.
Because here is what those buildings are. They are windowless boxes full of somebody else's intelligence, drinking power off our ground, sited here precisely because we have the transmission lines and the land and the water and a hundred years of history saying yes.
The power goes in. The intelligence belongs to a company headquartered somewhere else. The property tax stays, the construction jobs are real and temporary, and the permanent headcount in a hyperscale facility is a rounding error against what a plant employs.
And every megawatt those campuses reserve is a megawatt you do not get, at a price you did not set, drawn from a grid that has now missed its own reliability target two auctions running.
Energy out. Ownership elsewhere.
Second verse, same as the first.
I am not against the data centers. That is not the argument and anyone who tells you it is has not read this far. The buildings are coming and some of that is genuinely good for the tax base.
The argument is this. A region can host the AI economy without ever joining it. We have done that before, with a different fuel, and we know exactly what it costs, because we are still living in the answer.
The difference this time is that joining is actually available. In 1901 there was no version of this where an anthracite town owned a piece of the industrial revolution it was fueling. The capital did not exist here, the capability did not exist here, and nobody was offering.
In 2026 the capital is in the tax code, the capability ships to anyone with a credit card, and the only missing input is the decision.
That is new. In the entire history of this valley, that is new.
If you are not in Pennsylvania, do not put this down. Find your version, because you have one.
Appalachia sent out coal. The Permian sends out oil and gas. The Iron Range sent out ore. The Tennessee Valley sent out hydroelectric power that built aluminum smelters for someone else's war. Every one of those places supplied the energy for a transformation and watched the ownership settle somewhere with a better airport.
The pattern is not about coal. It is about who owns the thing the energy produces.
Right now the thing the energy produces is intelligence. And for the first time in the history of any of these places, the intelligence is purchasable by the people standing on the ground that powers it.
I want to close on the part that matters more than any number in this document.
Every one of us spent childhood waiting to be a grown-up.
You could not wait. You wanted the car, the money, the say-so, the right to walk into a room and be the one who decides. And then you got there, and somewhere in your forties it landed on you that being a kid was the whole thing. Not because it was easy. Because it was open. A Saturday morning had no agenda in it. You could build whatever you wanted out of whatever was lying around, and nobody had told you yet which things were impossible.
You started your business chasing that feeling. Whether you knew it at the time or not.
Ask a founder what they miss about the early years and almost nobody says the money. They say something harder to name. They say they miss being able to move.
In year three you could try something on a Tuesday. You had an idea at lunch, you walked out to the floor, and by Thursday you knew whether it worked. There was no committee. There was no system of record that had to be updated. There was no six-month implementation. If it was wrong you found out on Thursday and you tried the next thing on Friday, and that was not chaos. That was the most alive you have ever been at work.
Somewhere between then and now, that stopped.
And the reason is not that you lost your nerve. Everyone gets that story wrong, including a lot of the people telling it. You did not become cautious. You became committed. You built something that worked, and everything that worked hardened into a commitment: to a process, to a system, to a customer promise, to a way the plant runs on Tuesdays. Each one was correct when you made it. Every single one.
And then one morning you looked up and understood that you no longer run the business you built. You defend it.
There is a name for that condition. Scaling Trapped™: the business grew into a structure that now consumes the person who built it.
That is what happened. Not a loss of appetite. A loss of the thing that made you an entrepreneur in the first place, which was never courage and was never money.
That is what is being handed back to you. Not efficiency. Not headcount reduction. Not a chatbot on your website.
The ability to begin again. To dream something on a Tuesday and build it by Friday. To wonder what would happen if, and then find out, instead of filing the thought where you file all the others.
Think about what that means in your hands, because almost everyone describing this moment aims far too low.
You have wanted to know for six years whether that second line would work. Not a study. Not a consultant's deck. Know. You have had a plant layout in your head since 2019 that you never priced, because pricing it would have burned a quarter. You have watched a competitor's customer get treated badly and thought, we could serve that better, and then you looked at what finding out would cost and you let it go.
Those are not fantasies. Those are things you already decided were beyond reach, filed under later, and stopped bringing up in meetings.
Every one of them is now a Tuesday.
Watch the people who move from one industry into a completely different one and build something serious in both. They are not smarter than you. Many of them are not better operators than you. What they have is one thing: they never lost the ability to start over, so starting over never felt like betting the company. It felt like Saturday morning.
That capacity is not a personality trait. It is a structural condition. And it can be built.
That is what the architecture does. The question in your leadership meetings stops being can we afford to find out and becomes what should we find out first. Those are not two versions of the same meeting. They are two different companies, and only one of them is going to be worth running.
And once you can find out, you can build. Once you can build, you can compound. Compounding is what builds an Unassailable Moat, and that is where the wealth is. Not in the cost you took out. In the things you can now attempt that were priced out of reach for your entire career.
Now the second half, and this is the part I care about most.
When AI takes over the work that was only ever machine-substitute labor, the transcription and the reconciliation and the report nobody reads and the fourth time this week somebody retyped the same number into a different system, that work does not vanish into nothing. It vacates a seat.
And the person in that seat is still there.
What is left for them is the work that actually requires a human being. Judgment. Reading a customer. Knowing that the line sounds wrong before the sensor says so. Teaching the twenty-four-year-old what forty years of pattern recognition looks like from the inside. Deciding what should be built rather than counting what was.
The proportion of distinctly human work in your business goes up. Measurably. That is not a consolation prize offered to make automation palatable. It is the actual mechanism, and it is the reason I do this work instead of something easier.
I call it Raising the Humanity. Your people do not become less necessary. They become harder to replace, which is the only real job security anyone has ever had.
And here is what nobody tells you about that. It is not only your capacity to begin that comes back. It is theirs.
The maintenance lead who has had an idea about that line for two years and never had a way to test it. The scheduler who knows exactly what is wrong with the way you sequence and has never had the standing to prove it. The woman in quality who has been right about three things in a row and had to wait eighteen months each time to be believed.
Give a building full of people the ability to try something on a Tuesday and then stand back and watch what that does.
You never wanted to be the only one who gets to build. You wanted to build a place where building is what everybody does.
That is what a plant looks like on the other side of this. Not emptier. More alive. Running with the responsiveness you had in year three and the scale you have now, which is a combination that did not exist on any earlier surface, in any prior transition, at any point in the industrial history of this country.
And here is the part that should keep you up in the good way. Almost nobody has done it. Eighty percent of the field is standing exactly where you are standing, waiting for somebody to explain what any of this means.
The rails are going down. They are going down whether or not you are on the platform.
The only question left is whether the freight is yours.
I did not write these. Jim Rohn did, and he put them to millions of people across forty years, and I have never found four better ones.
Why did you do any of this in the first place.
Not the version you give the bank. The real one. You wanted to build something that was yours and make it good and have it matter to the people who work in it and the town it sits in. You wanted your name on something that was excellent.
That reason did not expire. It got buried under twenty years of operating. Everything in this document is about whether it gets dug back up while there is still a window to act in.
Why? Because you never stopped wanting it. You stopped believing it was still available.
This is the question that does the work, because this is where every reason lives that has kept you where you are.
Why not build the plant you have been drawing in your head since 2019? Why not find out about the second line? Why not go after the customer you know is being served badly? Why not be the company in your region that figures this out first instead of the one that reads about it?
Sit with the answers that came back. Because the honest ones sound like this: it costs too much. It takes too long. We do not have the people. Now is not the time.
Every one of those was true. That is what makes them so hard to argue with.
And every one of them is a statement about a world where the capital was not appropriated, the capability did not exist yet, and the clock was not running. Look at what is on the page in front of you. The building is expensable in one year, once, if you start before 2029. The capability ships today. The people constraint is the one your entire industry ranks first and nobody is architecting.
The reasons did not get answered by an argument. They got answered by the calendar.
So ask it again, and this time notice that you are answering out of a world that no longer exists.
Why not?
Somebody in your market is going to do this in the next thirty-six months.
They are not going to be smarter than you. I have spent forty-two years in rooms with people who moved first and people who moved late, and the movers were almost never the most intelligent people in the building. They were the ones who understood, earlier than everyone else, which surface they were standing on.
You have the plant. You have the customers. You have the people. You have forty years of knowing why the line sounds wrong before the sensor says so, which is precisely the knowledge this technology is useless without.
You are the one with everything that cannot be bought. The only thing you are missing is the decision, and that is the one thing nobody can sell you.
So when it happens in your market, and it will, and you are sitting there reading about it: what will you have been waiting for?
Why not you?
Because now is the only one of these that has a date on it.
Construction has to begin before January 1, 2029, which means the decision gets made in 2027, which starts in five months. The capacity your grid needed is being bought right now by somebody else. The people you will need in 2030 are being trained, or not, this year.
There is no version of this where you get to run it later on the same terms. The window opened because three things converged that have never converged before in my working life, and windows that open that wide close the same way.
You have been waiting for the right time. This is what the right time looks like from the inside. It never announces itself. It only ever shows up as a set of conditions that will not hold.
Why not now?
There is no good answer. There is only the one you will have to live with.
Here is what I do, stated plainly.
Everybody is selling AI tools. That is tactics. I work with you on the strategy and the architecture that runs it, so AI becomes your competitive advantage instead of an expense. And I can tell you, in dollars, what every month you wait is costing you.
This document showed you three clocks. It did not tell you where you sit against them, because I do not know your business and I will not pretend I do from a distance.
That is what the Strategic AI Intelligence Diagnostic™ is for. It is a working session, one on one, $750. We examine your actual operating reality: where your intelligence flows and where it leaks, what you are forfeiting in dollar terms right now, and what the first ninety days would have to look like. You leave with the number and the sequence whether or not you ever work with me again.
If you want it, the reach is direct.
No form. No calendar link. Write to me the way you would write to a peer, tell me what you make and roughly what you run, and I will tell you honestly whether a Diagnostic is worth your money.
If you would rather start further back, The Decision Room: Strategic AI for the Mid-Market Leader is $500 and includes a second seat, because this is not a decision anyone should make alone. Bring the person you would tell at the kitchen table.
And if none of that is for you, take the three clocks to your CPA and your lender this month anyway. That would be a good outcome. The point of this document was never to sell you something. It was to make sure that whatever you decide, you decided it, instead of arriving at it by default in 2029 and calling it bad luck.
Every figure above, in order of appearance. Canon codes in brackets for internal traceability.
Copyright © 2026 Carl J. Peterlin Jr. | Peterlin Intellectual Properties, LLC. All rights reserved.
The Rails Are Going Down Again, Edition One — first published August 2026 at carlpeterlin.com/rails-2026 by Carl J. Peterlin Jr., Strategic AI Intelligence Architect™. This edition is superseded by successor editions and is never revised. Figures were verified as of August 2026 and decay; see the sources section.
No portion of this doctrine may be reproduced, adopted as institutional policy, or distributed under another organization's name without written attribution to Carl J. Peterlin Jr. as the originating architect. Sharing with full attribution is encouraged.