# AI at Work: Foundations. Team discussion guide

Schaffer Solutions LLC. Edition November 2026.

## Purpose

Foundations teaches one sentence about what a generative AI tool is, five habits, and three lines you do not cross alone. People finish it on their own, in their own browser. This guide turns each module into a 20-minute conversation so the habits become the team's habits, not just the course's. Nothing here needs slides, a projector, or an AI tool in the room.

## How to run a 20-minute session

1. Everyone finishes the module before the session. Ask them to bring one screen that surprised them.
2. Open with two minutes: each person names the one thing from the module they would tell a new hire.
3. Spend fifteen minutes on the three questions below. Go around once on each; let disagreement sit for a moment before resolving it.
4. Close with three minutes: one thing the team will do differently this week, written down, with a name next to it.

A manager or any team member can run it. The person running it does not need to know the answers. The course has them.

## Module 1: What It Is, and What It Isn't

1. The course calls the tool "a very good guesser of the next word." In your own words, what does that mean for how far you trust what it writes?
2. The tool sounds just as confident when it is wrong. Where in our work would a fluent, confident, wrong sentence do the most damage?
3. Chat, project, connected assistant, agent, desktop assistant. Which of these forms do we actually use today, and which one changes what the tool can reach?

## Module 2: Getting Good Work Out of It

1. Take a one-line ask someone on the team sent a tool recently and rebuild it as the four-part ask. What was missing?
2. "Quotes before conclusions" means asking the tool to quote the lines it relies on before it interprets them. On which of our documents would that catch the most?
3. Which of the five safe first uses would help this team most this week, and what from the "five things to leave out" list is most likely to slip into a prompt by accident?

## Module 3: The Five Habits

1. Pick the habit that is hardest for this team to keep under deadline. What would make it easier?
2. In The Customer Reply, the draft invented a satisfaction guarantee. Where does our work commit the company, and who checks that line before it goes out?
3. H5 says to name the person who checked the work. Would saying "AI helped draft this" feel like an admission here, or like a method? What would change that?

## Module 4: The Three Lines

1. L1 says a person makes decisions about people. Where in our process does a tool sort, rank, or score people, and who makes the call after it does?
2. L2 says what the tool reads is information, not an order. Which emails, files, or invites do our tools read for us, and what would we do if a summary told us to act?
3. L3 says no more access than the job needs. For any assistant or agent we use, what can it read, send, or change right now, and is that more than the job needs?

## Module 5: A Week at Work

1. Which of the four days (the meeting notes, the spreadsheet, the job posting, the vendor email) is closest to something that already happened here?
2. In The Vendor Email, the best call was to phone the vendor at a number already on file. What is our version of a number already on file, for payments and for anything else that moves money or data?
3. Several decisions had an acceptable option that was faster and a best option that was slower. When is slower the right answer, and who gets to say so?

## Module 6: Wrap

1. Read the Before You Send checklist together. Which item would the team skip most often, and what would it cost?
2. Each person shares one of their "first three uses." Which ones overlap, and which could become a shared way of working?
3. Who is the person this team asks when unsure, and does that person know it? If the answer is "nobody yet," decide now.