Before we head into another week, here's something I've been thinking about. We spend a lot of time looking for the perfect AI prompt. But sometimes the prompt isn't the problem. AI may not know that you teach 150 students. Or that you're working in Blackboard. Or that your students are dual credit. Or that you need something accessible, practical and ready before Tuesday morning. So before asking AI to solve something this week, try this: “Before you help me, what do you need to know about my situation to give me a useful answer?” Then make it ask you one question at a time. I'm curious: What's one thing AI would need to know about your work to actually be helpful? I am going to be doing a presentation at an upcoming summit on this but would also like to hold a class for our members. What day of the week and times are best for you?

Before opening an AI tool this week, start somewhere else: identify one place where the work feels harder than it should. Maybe students keep asking the same question. Maybe faculty are spending too much time formatting materials. Maybe a training process requires the same preparation every time. That friction is more useful than a list of AI features. Once you know the problem, you can decide whether AI belongs in the solution at all. One Free Takeaway: Use this simple sequence: Notice the friction. Define the desired improvement. Then decide whether AI, a template, a process change, or no change is the best response. What is one small friction point in your work you would like to make easier this week? One AI Prompt: I am experiencing this recurring friction in my work: [DESCRIBE IT]. Help me identify the underlying task or process problem first. Then tell me whether AI could reasonably assist, whether a simpler non-AI solution might work better, and what should remain under human judgment. I am trying to build this skool and will be doing classes soon. What can I do to help you and how can I get you to start discussing topics on here?

"AI makes this faster" is not enough information to decide whether a workflow is better. Faster for the instructor can sometimes mean more confusing for the student. Easier for a department can create another step for faculty. More automated can also make a process less accessible or harder to explain. When evaluating an AI workflow, consider everyone affected by it, not only the person operating the tool. Good implementation should reduce unnecessary friction without quietly moving that friction somewhere else. Before adopting an AI workflow, ask three questions: Who saves time? Who gains work? What changes for the person receiving the result? Have you encountered an "efficient" process that actually created more work for someone else? Todays prompt: Evaluate this proposed workflow from the perspectives of the educator, learner, support staff, and institution: [DESCRIBE WORKFLOW]. Identify where it reduces work, where it may shift work to someone else, and any accessibility, privacy, clarity, or human-oversight concerns.

This week, notice what you are doing for the second, third, or tenth time. Repeated work is often where a small system can make the biggest difference. Instead of starting from scratch, look for something worth turning into a reusable starting point: an email framework, feedback template, course announcement, meeting agenda, or planning prompt. You still bring the judgment and context. The system simply gives you somewhere better to begin. Try the Repeat-to-Resource check: If you have created something three times, ask whether the useful structure can become a reusable template. (Parent or Student emails, creating a lab worksheet or even just putting together an agenda) What is something you keep recreating that probably deserves a template? Try this prompt in the AI of your choice: Review these examples of something I create repeatedly: [PASTE EXAMPLES]. Identify the consistent structure and create a reusable template with placeholders. Preserve the parts that require my professional judgment instead of automating them.

A little Friday thought experiment: imagine AI has officially joined your department or team meeting. It gets exactly one job, and you are not allowed to give it anything involving professional judgment or final decisions. What job are you handing over immediately? Taking notes? Organizing action items? Turning the discussion into a first-draft follow-up? Sorting ideas into categories? This is playful, but it also reveals where we see low-risk opportunities to reduce administrative work. One Free Takeaway: Look for tasks involving organizing, formatting, summarizing your own notes, or preparing first drafts before looking for ways to automate decisions. AI gets one job at your next meeting. What are you assigning it?