How to Teach AI Without Putting People to Sleep

## Why Does Most AI Training Fail? I've sat through enough bad tech trainings to know exactly what kills them. Someone stands at the front of the room, shares their screen, types a prompt into ChatGPT, and says "see how cool that is?" For an hour. Maybe two. The audience nods politely. A few people take notes. Everyone goes back to their desks and does exactly what they were doing before. **The training didn't fail because AI isn't useful. It failed because watching someone else use AI teaches you almost nothing.** ## What's Wrong with the Standard Approach? Most AI training makes three fundamental mistakes: ### 1. It's a Demo, Not Training There's a difference between showing someone what AI can do and teaching them how to use it. Demos are impressive. Training is practical. If the audience never touches a keyboard during your session, you gave a demo. ### 2. It Uses Generic Examples "Let's ask AI to write a poem!" Great — except nobody in the room writes poems for a living. When the examples don't match people's actual work, the connection between "this is cool" and "I should use this tomorrow" never happens. ### 3. It Assumes a Baseline That Doesn't Exist Most trainers assume everyone knows the basics. They don't. In every workshop I've run, there are people who have never opened ChatGPT. Not because they're behind — because nobody ever showed them. If you skip the basics, you lose half the room in the first five minutes. ## What Actually Works? After running AI workshops across manufacturing, insurance, hospitality, F&B, and service industries — for teams ranging from factory floor managers to C-suite executives — here's what I've found actually changes behavior. ### Start with the Basics (Yes, Really) I don't care if half the room uses AI daily. Start with the basics anyway. The daily users will pick up nuances they missed. The beginners won't feel left behind. And everyone starts from the same page. The basics aren't "what is a large language model." The basics are: here's where to type, here's how to talk to it, here's what to expect back. ### Lots of Practice This is the single most important thing. People need to type their own prompts, see their own results, and troubleshoot their own problems — during the session, not after. I structure my workshops so that for every 10 minutes of me talking, there are 15-20 minutes of the group practicing. The ratio matters. More doing, less listening. ### Use Their Actual Work Before every workshop, I learn what the team actually does. What reports do they write? What emails do they send? What data do they analyze? Then I build the exercises around those tasks. When someone sees AI draft a version of the exact report they spent 2 hours on last week — and it took 3 minutes — something clicks. That's the moment adoption starts. {% callout type="tip" title="The Magic Moment" %} Every workshop has a moment where someone's eyes go wide. It's always the same: they asked AI to do something they thought was too specific, and it nailed it. That's the moment they go from skeptic to user. Your job as a trainer is to engineer that moment. {% /callout %} ### Make It Safe to Be Bad at It The first prompt anyone types is going to be bad. The result will be generic or wrong or weird. That's normal. But if people feel judged for a bad prompt, they'll stop trying. I tell every room the same thing: "Your first prompt is supposed to be bad. That's how this works. The skill isn't writing a perfect prompt — it's knowing how to fix a bad one." ### Mix Teaching Styles Not everyone learns the same way. Some people need to see it. Some need to hear the logic. Some need to do it with their hands. I mix all three — show a concept, explain why it works, then have everyone try it immediately. ## A Simple Workshop Structure That Works If you're thinking about running an AI session for your team, here's the structure I use: ### Block 1: Orientation (15 min) - What AI actually is (in plain language, not jargon) - What it's good at and what it's bad at - Quick survey: where is everyone starting from? ### Block 2: The Basics (30 min) - Open the tool together - Type a basic prompt together - See what happens, discuss the output - Practice: everyone tries 3 prompts on their own ### Block 3: Real Work (45 min) - Pick 2-3 tasks the team actually does - Walk through how to use AI for each one - Extended practice: everyone works on their own tasks - Share results and troubleshoot together ### Block 4: Level Up (30 min) - Advanced techniques: giving AI context, roles, and constraints - How to iterate when the first answer isn't right - Building prompts for their specific recurring tasks - Q&A {% callout type="warning" title="The Follow-Up Matters" %} The workshop isn't the finish line — it's the starting line. Without follow-up within 1-2 weeks, most people revert to old habits. A quick check-in, a shared prompt library, or even a group chat where people share what's working can make the difference between a one-time event and lasting change. {% /callout %} ## The Goal Isn't AI Expertise Here's what I tell every group at the start: the goal today isn't to make you an AI expert. The goal is to make you comfortable enough to use it tomorrow. And the day after. And the day after that. Expertise comes from repetition. Comfort comes from a single good experience. My job is to give you that first good experience — the rest is practice. **The best AI training doesn't feel like training. It feels like someone finally showing you the shortcut you've been missing.** {% callout type="info" title="Need a Workshop for Your Team?" %} AI Ascend runs practical, hands-on AI workshops built around your team's actual work. No jargon, no boring lectures, no generic demos. Just real training that gets real people using AI the next day. Get in touch if your team is ready. {% /callout %}

How to Teach AI Without Putting People to Sleep

Let's Talk