The scarce asset is the objective function

AI is making execution cheaper faster than most businesses are making judgment clearer.

The scarce asset has moved.

In a labor-constrained company, the bottleneck often looks like capacity. There are not enough hours, not enough follow-up, not enough analysis, not enough people to keep every promise with the same care. New software looks valuable because it removes drag from execution.

Emad Mostaque’s Chapter 15 claim in The Last Economy is more useful than that. Once intelligent action becomes cheap, the constraint moves up a layer. The scarce thing is no longer mainly the ability to do more work. It is the ability to specify what the work is for.

Owners feel that shift first.

An agent can optimize appointment fill, response time, estimate conversion, or collections speed. It can run purchasing tighter, schedule crews harder, and cut support cost faster than a human manager used to. If the target is shallow, the system will still produce a shallow win. A clinic can improve collections by making billing more aggressive. A service company can improve utilization by overloading the best technician. A sales team can raise close rate by approving weak-fit customers who create chaos later.

The automation worked. The business got worse.

The operator contract comes first

Every serious automation path should start with a three-part contract.

First, name the optimization target. Not “help the team” or “save time.” State the score the system is meant to improve and the horizon that matters. Reduce days-to-close claims without increasing rework in the next thirty days. Improve schedule density without breaking on-time arrival or callback rate. Raise lead response speed without lowering booked-job quality.

Second, name the protected values. These are the things the system is not allowed to trade away in pursuit of the target. Customer trust. Operator dignity. Auditability. Apprentice teachability. Clinical safety. Cash accuracy. When those values stay implicit, the machine discovers them through damage.

Third, name the anti-goals. These are the unacceptable wins. Do not hide denials to improve dashboard cleanliness. Do not pressure customers into the wrong treatment plan to lift conversion. Do not turn the owner into the permanent exception handler just because the routing metric looks better. Anti-goals matter because optimization pressure always finds the exploit before management names it.

This is what alignment looks like in a real company. It is not a philosophy seminar about AI. It is a concrete operating discipline for owners who do not want to wake up inside a more efficient version of their own failure mode.

Hadto’s thesis depends on this

Hadto should not help businesses produce more output inside a vague objective. The better job is helping owners become more capable authors of the objective itself.

That is the difference between an employee with AI tools and an owner-operator with governing authority. The employee is asked to run faster inside a target someone else defined. An owner-operator has to decide what counts as a good win, what must remain protected, and which shortcuts would poison the business even if they look efficient this quarter.

Cheap intelligence raises the value of that role.

A future owner is not the person who touches every task. It is the person who can set the aim, define the constraints, inspect the evidence, and improve the system without surrendering judgment to the dashboard. If AI is going to expand human capability instead of shrinking people into button-pushers, more operators need that authority.

A business does not become more modern when it adds agents. It becomes more governable when the owner can say, in plain language, what the agents are for and what they are never allowed to become.

---

Source evidence used in this note: Emad Mostaque, The Last Economy, Chapter 15, “The Alignment Economy”, public PDF reviewed 2026-05-18. Hadto interpretation: owner objective functions as operating authority.