The work meeting is over. Now you need to explain what happened, brief leadership, and make sure the work gets done. Those are three different needs—and you can use the same transcript for all three. For this exercise, use the attached fictional bakery meeting transcript. Five employees discuss how to improve morning pickup orders. 👇 Open a new chat, upload the transcript, and try these prompts one at a time in the same chat. 1️⃣ Meeting Summary — What happened? Prompt 1: Create a meeting summary from this transcript for the attendees. Include their names and titles, the main problems discussed, agreed process improvements, decisions, and unresolved questions. Separate approved decisions from suggestions and proposals. Include timestamps for key decisions. Use only the transcript and flag uncertainty instead of guessing. 2️⃣ Executive Summary — What does leadership need to know? Prompt 2: Using the same transcript, write an executive summary of no more than 200 words for the bakery owner. Focus on the business problem, the agreed pilot, its scope and timing, proposed success measures, and risks or decisions still requiring attention. Distinguish proposed targets from confirmed commitments. Use only the transcript. Don’t claim that improvements have already occurred. 3️⃣ Task List — Who does what, by when? Prompt 3: Extract the agreed tasks from the transcript into a table with these columns: Task, Owner, Due Date or Timing, Dependencies or Approvals, and Source Timestamp. Include supporting contributors where stated. Convert relative dates using the meeting date of October 12, 2026. If an owner or deadline isn’t stated, write “Not specified.” List ongoing pilot responsibilities separately from one-time tasks. Don’t turn unapproved suggestions into assigned work. Now review the results: 📋 Does the meeting summary capture the discussion and decisions? 🎯 Does the executive summary give leadership the essentials? ✅ Does the task list accurately identify owners, deadlines, and approvals?

You don’t always need a detailed prompt to discover what AI can do. Try it with the photos, blueprint, and screenshot of a poem attached to this post. 1. Open a new chat in a LLM that accepts images. 2. Upload the first image and type: “Describe what you see.” 3. Upload the next image in the same chat. Then the next. Read what it picks up: objects, colors, visible text, layout, and details you might have overlooked. Compare its description with the image—did it miss anything or make an assumption? 📸 Now upload your own photo or photos. Try a screenshot, a handwritten note, an item you want to sell, or something you’re curious about. 💬 Share what surprised you?

AI can make an image look good. But does it speak to the right audience, include the right message, and represent the organization accurately? Let’s put it to the test. Same AI tool. Two different prompts. For this exercise, we’ll create a mentor recruitment graphic using Winning Futures’ volunteer information. Use the same two-round experiment as Lessons 1 and 2. Open a new chat for each round, then compare your results. 💡 Of course, you can change the prompts to include your own company, website URL, and logo. Try it and see what you get! Round 1: Simple Request Open a new chat in an AI tool that can create images and enter: "Create a social media image encouraging people to become mentors for Winning Futures. Use https://winningfutures.org/ for reference." 🔎 Review the image. Round 2: Clear Assignment Open another new chat in the same AI tool, then copy and paste this prompt. Modify it as appropriate: Begin Prompt -------| "Create a professional square social media graphic encouraging Metro Detroit business professionals to become volunteer Career Mentors with Winning Futures. Use https://winningfutures.org/ and its Be a Mentor page for accurate information and visual branding. Show an adult mentor collaborating with high-school students on career goals. Use a polished editorial illustration style, a clean layout, generous open space, and text that’s easy to read on a phone. Include this text: Help Shape a Future Become a Winning Futures Career Mentor Learn more at WinningFutures.org Use the official Winning Futures logo unchanged and match its brand colors. Do not redraw or invent the logo. If you can’t retrieve it, ask me to upload it before creating the final graphic. Don’t invent event details or statistics. 📎 Have the official Winning Futures logo file ready to upload if your AI tool asks for it. Access to a website doesn’t always mean the tool can retrieve its logo."

“Analyze this budget” sounds straightforward. But does AI know what you need to learn—or what your boss will want to see? In this exercise, you’ll give the same spreadsheet to the same LLM tool twice (pick your favorite ChatGPT, Claude, etc.). First, we will make a simple request. Then, we'll give it a clear assignment with specific calculations, evidence, and a report format. 👇 Try both rounds, then judge the results for yourself. Round 1: Simple Prompt Request Open a new chat, upload the sample spreadsheet, and enter this prompt: "Analyze this budget." 🔎 Review the response. Round 2: Clear, Concise Prompt Assignment Open another new chat in the same AI tool. Upload the spreadsheet again, then copy and paste this entire prompt: "Act as a nonprofit budget analyst. Review every tab in the attached workbook for January–June 2026. Create a professional, downloadable Word report I can share with my boss, no more than two pages. Include: 1. An executive summary of the overall financial position 2. The largest revenue shortfalls and expense variances, in dollars and percentages 3. Monthly trends that the six-month totals might hide 4. A comparison of program spending, participation visits, and cost per visit 5. Three recommended next steps, with questions we should investigate before making decisions Support each key finding with figures and the relevant worksheet and line item. Don’t double-count summary tables and underlying detail. Explain how you calculate variances. Use only the workbook, distinguish facts from possible explanations, and flag missing information. Visits are not unique students, and cost per visit does not establish program quality." 🔎 Now compare the two results. Which found more useful information? 🧮 Which made its figures and conclusions easier to verify? 📄 Which produced something you could share with your boss, board, or finance team? Think of it like assigning work to a coworker. “Analyze this budget” leaves a lot for them to figure out. A clear assignment explains what matters, how to support the findings, and what the finished work should look like.

🤔 How much does your prompt affect what AI gives you? Let’s put it to the test. Same documents. Same LLM. Two different prompts. You don’t need technical language or a secret formula. In this exercise, you’ll see what changes when you tell AI exactly what you need—and how you plan to use it. 👇 Try both rounds, then judge the results for yourself. 1️⃣ Prompt: Keep it simple Open a new chat in your favorite LLM, upload Documents A and B (attached), and type just one word: "Compare" Review the response. 2️⃣ Prompt: Give clear, concise directions Open another new chat in the same LLM. Upload both documents again, then copy and paste this prompt: “Compare Document A (original) with Document B (revised). Create a professional, clearly formatted Word comparison document that I can download and share with my boss. Include a brief executive summary followed by a comparison table. For each difference, identify the section, quote the exact old and new text, label it added, removed, or changed, and explain the effect on meaning. Use only the uploaded sources. Flag uncertainty instead of guessing.” Now compare the two results. Which caught more differences? Which was easier to verify? Which gave you something you could actually use? Try switching model and/or thinking settings up and down? Think of it like assigning work to a coworker. How much does the result depend on how clearly you explain the task? Share your observations in the comments.