Hook
If you are reading this, you are probably not new to the idea of AI. You have heard the talks, watched a peer demo something on their phone, maybe opened ChatGPT or Claude once, typed a question, got a generic answer and quietly closed the tab. Months later the unease has not gone away. People with your title are pulling ahead, and you are still not sure where to actually start.
Here is the reframe that matters. The distance between you and the executive who uses AI fluently is not years of study. It is one focused hour spent on real work, done correctly. Not a course. Not a certification. One hour that turns AI from a thing you have heard about into a thing you have actually used on your own desk.
This lesson is that hour, laid out step by step. By the end of it you will have set up your account properly, run AI on a real piece of your own work, learned the one prompting pattern that separates a useful answer from a useless one, and saved it so the second time is faster than the first. That is the whole goal: not mastery, the start of a habit.
Context
Most executives who stall with AI fail for the same three reasons, and none of them is about intelligence.
They treat it like Google. They type a short question, get a short generic answer and conclude there is nothing here. Claude is not a search engine. It is closer to a sharp, fast analyst who will do excellent work, but only if you brief it like you would brief a person. A one-line query gets a one-line-quality result.
They practice on fake tasks. They test it with a trivia question or "write me a poem," see a party trick and learn nothing that transfers to Monday. The only practice that builds fluency is practice on your actual work: the memo you owe the board, the policy you are drafting, the analysis sitting in your inbox.
They quit after one mediocre answer. The first response is rarely the best one. People who get value treat the first answer as a draft to push on, not a verdict on whether AI is worth their time.
Avoid those three and the first hour almost always lands. What follows is the protocol that gets you there. Use a real task while you read. Do not just read along, do it.
You feel yourself become a different person when you're supported by AI.
Steps
Step 1: Set up your account the right way (10 minutes)
Go to claude.ai and create an account. For real work, get the paid plan (Claude Pro). The free tier is fine for a first taste, but it limits how much you can do in a session, and nothing kills a new habit faster than hitting a wall mid-task. Treat the subscription the way you treat any tool that saves you an hour a week: trivially worth it.
Once you are in, notice one feature most people miss: Projects. A Project is a workspace where you can store context once (your role, your company, your style, key documents) so you do not have to re-explain yourself every time. You will use this in Step 5. For now, just know it is there.
Step 2: Bring one real piece of work, not a test
Pick something genuine from this week. The best first task is a piece of knowledge work that you find tedious but important. Strong candidates for an executive:
- A long email or report you need to summarize before a meeting
- A first draft you have been avoiding: a policy, a memo, a proposal
- A messy set of notes you need to turn into a structured document
- A decision where you want the strongest argument for and against
Choose one. Have the actual material in front of you. The realness is the entire point. You are not testing whether AI works. You are using it to move a real piece of work forward.
Step 3: Write the first prompt that actually works
This is the single most important skill in the hour, and it is simple. A good prompt has four parts: who you want the AI to be, the context it needs, the task, and the format you want back. Think of it as briefing a sharp new analyst who knows nothing about your situation yet.
Here is a reusable template. Copy it, fill in the brackets:
You are an experienced [role, e.g. chief of staff to a CHRO].
Context: [2 to 4 sentences. Who I am, what this is for,
who will read it, and anything that matters: tone,
constraints, the audience's seniority.]
Task: [exactly what you want done.]
Format: [how you want it back: a one-page memo, five
bullet points, a table, an email draft, etc.]
A worked example, the kind of thing an HR leader actually faces:
You are an experienced chief of staff to a CHRO at a
2,000-person company.
Context: My CEO asked for a short, calm point of view on
how we should approach AI in HR over the next year. The
audience is the executive team. They are skeptical of hype
and short on time. I want a measured, credible tone, not
breathless.
Task: Draft a one-page point of view with three priorities
and, for each, one concrete first step we could take this
quarter.
Format: A one-page memo with a short opening paragraph,
three numbered priorities, and a one-line first step under
each.
Paste your version in and send it. What comes back will be dramatically better than anything a one-line question produces, because you briefed it like a colleague instead of interrogating it like a search box.
Step 4: Push on the answer, do not accept the first draft
The first response is your starting point, not your output. This is where the real value lives, and it is just conversation. Try moves like these, in plain English:
- "Make the tone more direct and cut it by a third."
- "The second priority is weak. Give me three stronger alternatives."
- "Rewrite this for a board audience instead of the exec team."
- "What is the strongest objection someone could raise, and how would I answer it?"
Each follow-up sharpens the work. Two or three rounds of this and you will have something genuinely useful, often better than your unaided first draft and produced in a fraction of the time. Notice what is happening: you are not being replaced, you are directing. Your judgment about what is good is doing the work. The AI is just fast hands.
Step 5: Lock in the second use before you close the tab
A first use you never repeat is a party trick. A second use is the start of a habit. So before you move on, do one small thing to make next time easier.
If the task you just did is something you do regularly (and most executive work is), open a Project, give it a name, and paste a short description of your role, your company and how you like things written into its custom instructions. Drop in the prompt that worked. Next week, when the same kind of task lands, you start inside that Project and skip straight to the work. That is how an hour becomes a habit, and a habit becomes the fluency everyone else is talking about.
Recap
- The gap between you and a fluent AI user is not years of study. It is one focused hour on real work, done right.
- Three traps stall executives: treating AI like Google, practicing on fake tasks, and quitting after one mediocre answer. Avoid all three.
- Set up Claude properly, including the paid plan and Projects. Do not over-prepare. Understanding follows use.
- Brief the AI like a sharp new analyst: role, context, task, format. That one pattern is most of the skill.
- Push on the first answer two or three times. The value is in the iteration, and your judgment is what does the directing.
- Before you close the tab, save what worked into a Project so the second use is faster than the first. Reps build fluency.
You have now done the thing most people in your position are still only worrying about. The next lesson explains why this matters so much for your career, in dollars: the AI Wage Gap, what it is and which side of it you intend to be on.