Teaching purpose
Students should leave able to:
- distinguish a prediction machine from an agent;
- distinguish artifact form from artifact authorship;
- place Actual, Predicted, and Ideal correctly in the Reality Equation;
- explain why editing changes the numerator;
- state the exact absorption case: Reality equals one, ln(1) = 0, and no human attention is stolen.
Materials
- A board or projected blank page.
- One printed Monday report, real or invented, placed face down.
- Optional slide showing a three-dimensional probability mountain.
Do not begin with slides about AI. Begin with breathing.
Opening: Make attention visible — 0 to 7 minutes
Spoken script
Before we talk about AI, I want you to notice your breathing.
Do not change it. Just notice it.
Most of you were breathing perfectly well before I mentioned it. But you were not paying attention to it.
Why?
Breathing is not unimportant. If it stops, nothing else in this room matters.
Breathing disappeared from attention because it was predicted well enough that consciousness did not need to enter the loop.
Now imagine that your breath suddenly catches.
Your attention moves immediately.
The importance of breathing did not change. The prediction changed. Something that was expected became surprising.
Surprise stole attention.
That difference gives us the subject of today’s lecture.
Absorption is not the disappearance of function.
Absorption is the disappearance of stolen attention.
Ask the room
What else in your life matters but ordinarily disappears from attention because it behaves as predicted?
Let students offer examples: heartbeat, balance, a refrigerator, a car starting, a phone unlocking, a familiar route.
For each example, ask what would steal attention back.
Board
Write:
Surprise steals attention.
Then write:
Human attention is normalized accumulated surprise.
Do not formalize the equation yet.
Prediction is not action — 7 to 15 minutes
Spoken script
Now we can approach AI without beginning with hype.
AI is not first a worker.
AI is first a synthetic prediction machine.
That sentence is easy to say and difficult to preserve, because almost every business word pushes us back toward work.
We call AI an employee.
We give it a task.
We ask whether it completed the task correctly.
We build an approval queue and assign a person to supervise it.
We have taken a prediction machine and dressed it as a worker.
A worker works.
An agent acts.
A prediction machine predicts.
Those are different categories.
Suppose a prediction machine predicts an invoice.
It predicts the customer, line items, rate, taxes, date, and payment terms. The invoice may be perfectly formatted.
Has the prediction machine sent it?
No.
Has it charged the customer?
No.
Has it recorded a legal obligation?
No.
It predicted the invoice.
Sending, charging, and recording are actions. An agent, a person, a workflow, or an operating system acts.
Prediction and action can happen close together. They can sit behind one button. That does not make them the same event.
Board
Write three lines:
A prediction machine predicts artifacts.
>
An agent acts.
>
Prediction is not action.
Quick classification
Read each sentence. Ask the room to answer “prediction” or “action.”
1. A system produces the invoice it expects to exist next month. 2. A system emails the invoice to a customer. 3. A system produces the shipment it expects a household to need next week. 4. A warehouse releases the goods to a driver. 5. A system produces the book it expects a company should place on its website. 6. A person edits Chapter 4 and publishes the file.
Answers: prediction, action, prediction, action, prediction, action.
The artifact does not reveal the author — 15 to 25 minutes
Spoken script
Turn over the Monday report.
What is this?
You will probably say it is a report.
You are right about its form.
But you have not told me who authored its existence.
A person could have gathered the numbers, chosen what mattered, written the explanation, and attached the file.
Or a prediction machine could have predicted Monday morning’s report.
Both artifacts might have the same title.
Both might contain the same numbers.
Both might arrive in the same inbox.
The noun does not determine whether the artifact is real or predicted.
Authorship does.
Now imagine a flooring subcontractor here in Charleston.
The owner opens the CRM and finds a new project called Four Seasons Charleston.
It contains the predicted square footage, flooring types, technical specifications, drawings, timeline, likely general contractor, likely architect, likely competitors, project leaders, and suggestions for how to win.
But no RFP arrived.
No bid appeared on PlanHub.
No salesperson entered the lead.
Who authored the project?
The prediction machine.
The CRM record is the form.
The project is a predicted artifact.
This is what a prediction machine manufactures.
It manufactures predictions.
Board
Write:
The noun describes the form.
>
Authorship identifies the artifact.
Then list:
- predicted book;
- predicted report;
- predicted invoice;
- predicted purchase order;
- predicted CRM project;
- predicted shipment;
- predicted website;
- predicted lecture.
Discussion
Ask students to name one clue that might help a prediction machine predict the Four Seasons project.
Accept clues such as land transactions, planning documents, architectural relationships, hotel announcements, permits, procurement patterns, or similar developments.
Then make the important correction:
The clues are not the project. They are part of the pattern. The resolved project is the prediction.
The Reality Equation — 25 to 35 minutes
Spoken script
Now we can place the predicted artifact precisely.
We will use a book because the word is short and the artifact is easy to imagine.
There is an Actual Book.
There is a Predicted Book.
There is an Ideal Book.
The Actual Book is in the numerator. It is past tense. It was published.
The Expected Book is in the denominator.
Expectation is complex. It has a real component and an imaginary component.
The real component is the Predicted Book.
The imaginary component is the Ideal Book.
The Ideal Book is the book as perfect expression. It does not come from AI.
The Predicted Book comes from the prediction machine.
Picture a three-dimensional probability mountain. The landscape contains possible books. The peak is the highest resolved probability at that moment.
That peak is the Predicted Book.
The value represents the whole artifact: title, cover, table of contents, front matter, 473 pages, back matter, everything.
That amount of compression should not trouble us. The price of oil compresses geology, weather, war, shipping, currency, refining, storage, labor, regulation, and demand into one number.
A single value can represent enormous complexity.
Board
Write:
\[ R = \frac{A}{E} \]
Then:
\[ E = P + iI \]
Then the book form:
\[ R_{book} = \frac{A_{book}}{P_{book} + iI_{book}} \]
Label each term:
- Actual is the numerator.
- Predicted is the real component of the denominator.
- Ideal is the imaginary component of the denominator.
- Reality is the quotient.
Check for understanding
Ask:
Where does the published book belong?
Answer: Actual, in the numerator.
Where does the synthetic prediction machine belong?
Answer: It supplies the Predicted Book, the real component of Expectation in the denominator.
Where does the Ideal Book belong?
Answer: The imaginary component of Expectation in the denominator.
Does the synthetic prediction machine manufacture the Ideal Book?
Answer: No.
The absorption case — 35 to 43 minutes
Spoken script
Suppose the synthetic prediction machine predicts the entire book.
The book is published.
Nothing is corrected.
Nothing is amended.
Nothing is edited.
The Actual Book in the numerator is the Predicted Book in the denominator.
When the Ideal component is zero, the numerator and denominator match.
Reality equals one.
The natural logarithm of one is zero.
No human attention is stolen.
That is absorption.
Now change one sentence.
The prediction machine predicted one book. A different book was published.
The numerator changed.
The gap between Actual and Expectation is surprise.
Surprise stole attention.
The artifact was not absorbed.
Notice how precise this is.
We are not estimating how hard the editing felt.
We are not saying the editor was annoyed.
We are not asking whether the change took five seconds.
We are looking at the quotient.
Did the numerator and denominator match?
If they matched, Reality equals one. ln(1) = 0. No human attention is stolen.
If they did not match, the gap recruited attention.
Board
Write the canonical statement in full:
Absorption occurs when the predicted artifact and the actual artifact are the same.
>
The numerator and denominator match.
>
Reality equals one.
>
ln(1) = 0.
>
No human attention is stolen.
Pause after the final line.
Application exercise — 43 to 48 minutes
Ask students to work in pairs. Give them one of these artifacts:
- a weekly sales report;
- a purchase order;
- a college lecture;
- a customer proposal;
- a restaurant food order;
- a construction project in a CRM.
Each pair answers five questions:
1. What would the real business artifact look like? 2. What would the predicted artifact look like? 3. Who or what authors each one? 4. What action could follow the prediction? 5. What change would show that the prediction stole attention?
Invite two pairs to report back.
Correct category errors immediately but gently. In particular, do not let “the AI sent it” stand when students are trying to describe prediction. Ask whether they mean the prediction machine predicted the artifact or an agent acted upon it.
Closing — 48 to 50 minutes
Spoken script
We began with breathing.
Breathing matters. It disappears from attention because it is predicted well enough to produce no unresolved surprise.
AI matters for the same structural reason.
It gives economic life another prediction machine.
The machine can predict a complete artifact: a book, report, invoice, project, shipment, website, analysis, forecast, or lecture.
The artifact’s form does not reveal its author.
The prediction machine predicts.
An agent acts.
And when the Predicted artifact and the Actual artifact are the same, the numerator and denominator match.
Reality equals one.
ln(1) = 0.
No human attention is stolen.
That is absorption.