Last edition said the alignment is temporary. This edition is about the room where you decide what to build inside it.
Last Thursday I worked an exhibitor table at a manufacturing AI summit in Pennsylvania. I had ordered seventy-five copies of my two books for that table. They never arrived. What carried every conversation that day was four stapled pages of dollar figures: the federal capital already set aside for a manufacturer, line by line, with the law or the program beside every number.
Not one leader asked me what AI could do. Every one of them asked the same two questions in different words. What is it worth? And what does it cost me to wait? The person who stayed at my table longest was a CPA, the one person in that building paid to be unimpressed.
Nobody asked what AI could do. They asked what it was worth.
Those are your board's questions. Every AI decision in your business ends at one table: a board, a family, a banker, or the three people whose signatures it takes to spend real money. That table speaks one language, the language of capital. AI keeps showing up there speaking something else.
Picture the next time AI comes up at your table.
You hand around one page, the same page every machine on your floor once had to earn. What it costs. What is already allocated to pay for it. What it returns, and when. What every month of waiting costs, in dollars. And the point, written down in advance, where the table reads the evidence and decides whether to keep going.
Nobody in the room needs to understand a model. Your controller reads the capital line and stops frowning, because a qualifying manufacturer has $750,000 to $2,000,000 and more already allocated to it, in tax you keep and credit you can draw. Your CFO checks the timing: on $300,000 to $500,000 of qualifying equipment and off-the-shelf software, at a 30% effective rate, Section 179 moves $90,000 to $150,000 of tax savings into the first year. That is timing, not a gain, and saying so out loud earns more trust at that table than any demo ever has. Your banker sees a plan built in stages with a stop at the end of each one, which is the only kind of plan a banker has ever liked.
The vote is short. For the first time, the table was asked a question it knows how to answer.
Now look at how AI usually reaches that table.
It arrives as a demo somebody saw at a trade show. A vendor's slide with a percentage on it and no source. A pilot that purchasing started on a company card. Nobody can say what it returns, because nobody decided in advance what it was supposed to return.
So the table does one of two things. It says no. Or it approves something small enough not to matter, and the business ends up with Bolt-On AI: tools bolted onto a business that was never architected for them. A chatbot on the website. A subscription in sales that never talks to the schedule. A note-taker in the conference room that nobody reads. Each one renews every month, and none of them adds up to anything.
Your board has never approved a machine because the salesman's video was impressive. It should not approve AI that way either.
Your board already knows the difference between an expense and an asset. That difference is the whole AI decision, and almost nobody puts it in those terms.
Bought as tools, AI is an expense. It is a monthly charge on a card, and the day you stop paying, it is gone, along with everything it did for you. Your quotes, your fixes, your customers' reasons for staying went through somebody else's system, and none of it stays with you.
A resource that depreciates the moment you stop paying is a cost, not an asset.
Architected, AI builds something that stays. Every quote your estimator prices, every job that went sideways and the fix that saved it, every reason a customer came back, lands in one place your business owns. That is your Built Intelligence: everything your business knows that no competitor does. How it prices. How it makes the part right the first time. Who buys, why, and what keeps them. You paid for it in salaries, in scrap, and in bids you lost and learned from. It is not something your business has. It is your business.
Your newest machining center is worth less every year it runs. That is what the depreciation schedule is for. Built Intelligence runs the other way. It is worth more every time the business works, because every job teaches the next one.
Intelligence compounds. It does not amortize.
That is the version of your company worth bringing to the board: a company that becomes smarter every time it works, built on capital that was set aside for it before you asked.
If you have already said no to AI at your own table, you were not behind. You were doing your job.
The discipline that makes a machine earn its floor space is the same discipline that carried your business through the downturns that took competitors with them. Every vendor who walked in with a demo and no payback number deserved the answer you gave. Your question was right. AI kept showing up without an answer to it.
The instinct that built your business does not need to change. It needs the right page in front of it.
You put a payback number on every machine before it earns its floor space. You are not doing it with AI.
A careful table does not say yes to something large all at once. It says yes to a stage.
In Edition 21 I defined the gate: the decision, made before a dollar is deployed, about what evidence will tell you to continue, change course, or stop, and who gets to read it. Build in stages, each one built to pay for the next, each one ending at a gate your board reads. That is how a disciplined table approves something big without betting the business on it.
And if the table decides to wait, it should know what it is deciding. Tabling AI feels free. It is not. Every month, the capital set aside for you sits unclaimed, the know-how in your building stays in a few heads, and the competitor who brought the right page to his own board pulls further ahead.
Not yet is a decision. It is the only one that arrives at your board without a number.
I can tell you that number, in dollars, for every month you wait.
Most leaders can describe what AI might do for them. Very few can walk into their own board meeting with the page that prices it. The Strategic AI Intelligence Diagnostic™ is a working session that builds that page: the cost, the capital, the company, and your first move. Leaders walk out seeing a version of their business they could not see when they walked in. $750.
If your board is going to hear about AI this year, let it hear the number first: [email protected]
The Decision Room: Strategic AI for the Mid-Market Leader is where you decide what version of that company is worth building. Ten decisions. Made explicitly, at the leadership level, before a dollar is deployed. $500. Two seats: yours and one for the person who should be in this decision with you. carlpeterlin.com/decision-room