Editorial
The Inversion Question
Last week I dedicated this edition to the AI Money Stack and walked readers to three rungs operators are charging for AI services, the edition resonated well with professionals and operators looking to learn more about the AI business models that are shaping the way we conduct business and the opportunities that are opening up as AI compresses the value to knowledge, coordination and execution (See live on the subject) almost down to zero. Two weeks ago, I wrote about the Stack Audit, and gave readers some guidance and prompts to save on tokens, a growing concern as AI is becoming indispensable to conduct business. This week I am attending a request of some operators who are worried about the impact AI could have in their decision making and capacity to judge, discern what should remain human. In other words, “what should AI never touch in my work?”
Juan
THIS WEEK BY THE NUMBERS
5.4 percent. What the Federal Reserve measured on April 3 about how much work time AI is actually saving across the U.S. economy. Two hours and ten minutes a week on a 40-hour week. The most credible source you can cite in front of a client right now.
Nine hours. What sales decks have been promising for AI productivity, per the WRITER 2026 Enterprise AI Adoption Survey. The gap between the promise and what the Fed actually measured is four times. That gap is the math the person paying you is about to question.
1,200 employees. The size of the group studied in HBR’s May 2026 piece “The Psychological Costs of Adopting AI.” It measured six things that happen when people use AI day in and day out, from feeling like the decision is not really theirs to seeing their own skill fade. When people stop using AI, that is not a problem to fix. It is people sensing the line.
56 percent. The share of companies that have now formally named a person to own their AI agents, per Gartner and IDC’s Q2 2026 data. Up from 11 percent in 2024. The job is just starting to take shape. The question is who that named person is and what authority they actually have.
Story 1: The Cognitive Cost Data the Early Adopters Are Reading
What happened. Harvard Business Review published “The Psychological Costs of Adopting AI” in May 2026. The study followed 1,200 employees and measured what happens as AI use goes deep. People stop holding the problem in their head. The decision starts feeling like the tool’s, not theirs. Their skill atrophies. Their contribution gets harder to point at. Their role starts to feel less like theirs. The study found these effects make people stop using AI even when the tools work.
What it means. Most AI rollout content treats people pulling back from AI as a problem to fix. The HBR data flips that. The people pulling back are sensing the line where their judgment stops showing up in front of the people paying them. They are not wrong. They are the first wave of what every operator is going to feel in the next eighteen months.
What I am observing. The operators raising their fees right now are the ones who can name the line in their own work. The line is different per business. A fractional CFO has one line. A solo marketing consultant has another. The work is naming yours before a client asks.
Story 2: The Fed Number That Resets Every ROI Projection
What happened. On April 3, the Federal Reserve published the cleanest measure we have of how much work time AI is actually saving across the U.S. economy. The answer is 5.4 percent. Two hours and ten minutes a week on a 40-hour week. Sales decks have been quoting nine hours. The Fed number is four times lower than the best vendor case and barely above what the worst users save. ManpowerGroup’s 2026 Workforce Barometer found the same gap from the worker side, with people feeling less confident even as usage went up.
What it means. Any spreadsheet your client built on the nine-hour promise is sitting on a number four times too high. When the people signing checks rerun the math with the Fed number, the savings collapse. The operator ready to walk into that conversation with the right numbers wins the work. The operator still pitching the old vendor math walks out.
What I am observing. The flip side of the 5.4 percent number is that 94.6 percent of the work week stays human. That is the ground where the inversion question matters. The operators who can name what AI is not allowed to touch in that 94.6 percent are the ones charging the premium fee.
THE PLAYBOOK
How to audit your own scope creep this week. Four moves you can run before Friday.
Move 1: List what you are handing to AI. Open the AI tool you use most. Look at the last 30 days of prompts. Write down five tasks you let AI do for you in that time. The list is the audit.
Move 2: Mark the line. Next to each task, write one letter. K if you would keep letting AI do it (AI does this well, the output is fine). P if you should pull it back (this should have stayed with you). M if it is mixed (AI helps but you make the call). The letters tell you where the line is.
Move 3: Rewrite one client offer. Take one client or one offer you are running right now. Write the line out. “AI does X. I do Y. The decision on Z is mine, not the tool’s.” Show it to the client. The conversation that follows is your calibration data.
Move 4: Send one pitch with the line in it. Take the rung you picked last week from the Money Stack issue and add the line. “Here is what I deliver. Here is what AI handles in the delivery. Here is what I will not let AI touch in your project.” Send it to one buyer. What they say back, or do not say, tells you whether the line is right.
The point this week is calibration, not closing. Operators who can name the line they will not cross are the ones getting paid to hold it.
CLARITY
Three prompts. Each produces a deliverable you can use the same day. Run them in order to audit one piece of your AI use this week.
The Scope Audit Prompt. Reviews your last 30 days of AI use across one tool, identifies three tasks you delegated that should have stayed with you, and gives you a one-sentence reason for each. Run before any pricing conversation, any client check-in, or any week where you have lost track of where your judgment is showing up. Use the output to write the boundary line in your next proposal.
The Sign-Off Test. For any AI output you are about to ship under your name, asks three questions: does this represent my actual judgment, can I defend each claim if challenged, would I have written this myself given the time. Returns a go/no-go plus the specific edits needed to make the output yours. Run before sending anything to a client.
The Reverse Brief. Most prompts tell AI what you want it to do. This one tells AI what you do not want it to touch. Returns a clean output plus a flagged list of any boundary the AI would have crossed without the instruction. Use when delegating anything sensitive: client communications, pricing decisions, strategic framing, anything with reputation cost.
The Scope Audit tells you where your judgment has drifted. The Sign-Off Test catches it before it ships. The Reverse Brief prevents it on the way in. Three prompts, one boundary discipline.
THIS WEEK AT BUILD TO THRIVE
Let’s setup your AI Micro Consulting Business. Two friendly 1:1 sessions. Seven components. Built on the Offer Builder AI Coach system. You walk out with a positioned, priced, repeatable practice in your own voice.
My new segment. BUILD TO THRIVE WITH AI. Let me know what you think!
This week is all about the AI Micro Consultant.
Thank you for reading.
Juan
Awesome prompts as always mate 🥳
Naming what stays human actually makes the AI parts work better