Executives & Directors · Module 00 · Chapter 001

Welcome and How This Program Works

This is a program for people who decide about AI, not people who build it. Level 1 asks ten executive questions in a deliberate order, and every chapter has five parts, so you can read it, test yourself on it, or review it in two minutes before a meeting. It works only if you use it on a real decision.

≈ 12 min read

After this chapter you can

Explain what the program is for, and what it deliberately does not teach.
Describe the four levels and the eleven Level 1 modules, and the question each module answers.
Choose a route through Level 1 - the full sequence, a role track or the cheat sheet - and the format that fits each moment.
Use the five-step method, with practice testing, discussion and a dated action, to turn a chapter into a better decision.

Suppose it is a Monday and three AI items reach you before ten o’clock. A supplier offers to have an AI agent handle customer refund requests, and wants a pilot decision by Friday. A staff representative asks whether employees may paste customer emails into a public chatbot, because many already do. And the chair asks for “our AI strategy” at the next board meeting.

None of these is a technology question. Each is a decision about money, risk, people or direction, and each will be made by someone this week. The only open question is whether the person making it understands enough about AI to make it well: to know what to ask the supplier, what the staff question really exposes, and what a strategy needs to contain before it deserves the name.

That Monday is what this program is built for. It does not try to turn you into an engineer. It gives you the mental models, the evidence and the questions to lead decisions like these, and a way of studying that makes them stick.

What this program is

Much AI training starts from the technology and asks what can AI do? That question is useful, and the program answers it carefully. But it is not the question a leader is paid to answer. The leader’s question is different: what should our organization do differently because AI exists?

Figure 0.1.1 The program teaches the technology only as far as a leader’s decisions need it.

That choice shapes everything that follows. You will meet technical ideas, including how language models work, why they make confident mistakes and what an agent is, but always at the depth a decision needs. You will not write code, train a model, sit a vendor certification or learn the menus of a particular product. Tools change every few months; the program names them only as examples and builds on what lasts: how value is created and measured, how risk is judged, how money is spent and how organizations change.

The program also makes a promise about evidence. Every number in it has a named source with a date on which someone last checked it, and claims that could not be checked were removed or labeled as illustrations. Chapters on fast-moving topics carry a higher volatility rating and are re-checked more often. You should be able to repeat any figure in this program to a skeptical board and say where it came from.

Four levels, one language

The academy has four levels, each written for a different job.

Figure 0.1.2 Four levels, one vocabulary. Terms mean the same thing at every level.

Level 1, this one, is the first to be built; the other three will follow in the same format. The levels differ in depth, not in language: when a director says evaluation, grounding or agent, it means what it means to the engineer who builds the system. That shared vocabulary matters more than it sounds. The handover between the person who approves a project and the people who deliver it is where misunderstandings often cost the most, and a common language is a cheap protection against them.

Eleven modules, ten questions

Level 1 has 130 chapters in eleven modules. After this introductory module, each module answers one executive question, and the order is deliberate.

Figure 0.1.3 Level 1: 130 chapters in eleven modules. Each module answers one question before the next becomes useful.

The sequence builds a way of deciding before it hands you a list of opportunities. A leader who starts with fifty use cases will be excited, but cannot yet tell which of them matter, which are feasible, what they cost or what they put at risk. So the program first explains why AI matters and what it is, then how to judge value and make strategic choices, and only then where to apply it. Risk, governance and economics come before the roadmap, because a roadmap that ignores them will not survive its first review. The final module looks ahead, for preparedness rather than prediction.

Each concept has one home chapter, where it is taught in full. Other chapters recall it in a sentence and name that home, so you never meet the same idea three times.

One chapter, five parts

Every chapter is built from the same five parts, each useful at a different moment.

Figure 0.1.4 Read it, test yourself, or review it in two minutes. The content is the same in every part.

The chapter text is the fullest version: the argument, the evidence, common misconceptions, questions and answers, and the references. The figures can be stepped through, so you can follow a process or a comparison one stage at a time. The self-check quiz at the end lets you test what stayed with you, and the glossary defines the terms the chapter introduces. The cheat-sheet card compresses the chapter into its one idea, three key points, the terms it defines, a red flag to avoid and a line worth remembering.

Because every part comes from one source, they are built not to disagree. A number in a figure is the same number, with the same source, as the one in the text.

Three ways through

You do not have to read 130 chapters in order, although the program is designed to reward it. There are three practical routes.

Figure 0.1.5 Start with the route that fits your time, then fill in the rest.

The role tracks are curated reading lists of ten to fourteen chapters for the CEO, the CFO, the COO, risk and legal, and the CTO and CIO. A reading group can follow a track as a playlist, one chapter per week.

The cheat sheet gathers every chapter’s card on one page per level: a map of every chapter’s one idea, the frameworks to draw from memory, a glossary, red flags and interview-style questions with model answers. A set of chapters is marked as core, so the shortest pass through it fits in the time before a board meeting.

Testing yourself beats rereading

Each chapter can be read alone in about fifteen minutes. Two features of it, the quiz and the dated action, are there for a reason you can check. The quiz is not a test of you; it is a learning technique. In a well-known experiment, students read a prose passage and then either restudied it or tried to recall it. Five minutes later, the restudy group remembered more. One week later, the order had flipped: students who had practiced recall three times remembered 61 percent of the passage, and those who had only reread it remembered 40 percent. The rereaders had also been the most confident they would remember it¹.

Figure 0.1.6 Recall one week later. Testing yourself beats rereading, even though rereading feels more reassuring.

A large review of learning techniques reached the same verdict across many studies: practice testing and spacing study out over time are among the most effective techniques, while rereading and highlighting, habits many learners rely on, add little². That is why every chapter ends with a check, and why the cheat-sheet card exists: two minutes with a card a week later is spaced practice.

How to use it: understand, question, apply, discuss, act

The difference between finishing a course and learning from it is what you do between chapters. The program asks you to run each chapter through five steps.

Figure 0.1.7 The aim is not to finish the program but to decide better because of it.

Understand the idea on its own terms before testing it against your experience. Question it: when would it not hold, and what would make it hard in your organization? A framework that does not fit is useful too, because the reason it does not fit often reveals a missing capability. Apply it to a real case, such as a proposal on your desk or a process that frustrates your customers.

Discuss it with the people who see other parts of the problem. A technology leader sees feasibility, a finance leader sees economics, a lawyer sees exposure, and the people doing the work see why adoption will stall. This is more than courtesy. A meta-analysis of 89 studies found that a supportive work environment is linked to people actually using what they were trained to do, and that the link is stronger for open skills such as leadership than for closed skills such as operating software³. Leadership learning that stays in one head rarely changes an organization.

Act, and make the action specific. The action can be small: asking for a baseline, challenging an assumption, pausing a pilot. Give it a trigger and a date: “At Thursday’s steering meeting, I will ask how the supplier measured error rates.” Your AI Leadership Starting Point, the last chapter of this module, shows why that form works and uses it for your first commitment.

A running notebook makes the loop easy. Keep five headings: ideas worth keeping, open questions, opportunities, risks and actions. By the end of Level 1 it is a record of how your thinking changed, and the raw material for the roadmap you build in Module 09.

Story: a day with the program

What follows is an illustration, not a real case. It shows how the parts fit into an ordinary working day.

Figure 0.1.8 An illustration: five parts, one ordinary day, one decision made better.

The visitor operations director of a large public aquarium has a meeting at half past nine about a supplier’s proposal: an AI assistant to answer ticket, booking and refund questions. Over breakfast she reads the cheat-sheet card for What Does AI Value Actually Mean?. One line stays with her: value is a measured change against what would have happened without AI.

In the meeting, the supplier promises faster responses. She asks what response times are today, and nobody knows. The meeting ends with a request for a baseline instead of a signature.

At lunch she spends twenty minutes on the same idea with her guest-services, finance and IT leads. She asks each person to name the one measure that would prove the pilot worked. Finance names cost per resolved inquiry; the guest-services lead names complaints about refunds. They agree to track both.

At four she writes her action in the notebook: “Before the pilot decision on the 20th, we will have four weeks of baseline data on response time and refund complaints.” In the evening she reads the full book chapter, because she wants the evidence before she defends the request for patience to her chief executive.

Nothing in her day required technical expertise. It required the right question at the right moment, and a habit of turning what she read into a decision.

What this means for leaders

Treat this program as practice for decisions you already face, not as reading to finish. Choose the route that fits your time, protect a fixed slot for it, bring real cases to it, and involve the people who will carry out what you decide. The fastest way to get value is to start with the decision on your desk.

Check yourself

Level 1 teaches you to operate specific AI products.
Every chapter of Level 1 comes with a self-check quiz, step-through figures, glossary terms and a cheat-sheet card.
Rereading a chapter is the most reliable way to remember it.
The modules can be taken in any order without losing anything.
Discussing a chapter with your team helps you apply it.
Two minutes with a chapter’s cheat-sheet card a week later is a form of spaced practice.

Reflection: one decision, five lines

What comes next

You now know what the program is, how it is organized and how to use it. The harder question is why it deserves your time at all, when there are specialists who understand AI far better than any executive needs to. The next chapter, Why Executives Need AI Literacy, makes that case.

References

Henry L. Roediger III and Jeffrey D. Karpicke. Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science 17(3), 249-255. 2006.
John Dunlosky, Katherine A. Rawson, Elizabeth J. Marsh, Mitchell J. Nathan and Daniel T. Willingham. Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest 14(1), 4-58. 2013.
Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin and Jason L. Huang. Transfer of training: A meta-analytic review. Journal of Management 36(4), 1065-1105. 2010.