For experienced instructional designers who already use AI, but want more from it than faster content production. Use AI throughout your design process. You still make the decisions that require your experience.
Write down those decisions and corrections as you work. Project by project, they become your own design standard. So AI gets better at supporting the way you work.
But too much of your day goes into fixing AI output.
A scenario with obvious wrong answers. Forty AI-generated slides presented as a finished course. You can see what’s missing, and the design work still lands on your desk.
AI can help earlier, with the problem itself, and later, when you review the work. Your judgement belongs throughout the process.
AI doesn’t know the design rules you carry in your head.
Make those rules explicit. Record why you choose a scenario, change a question or reject an answer. Then the next project can start with what you’ve already learned.
Learning that was built fast, but where the decisions behind it are no longer clear.
Six months later, the course is still there, but nobody knows why that scenario was chosen or what problem it was meant to solve. The next project starts, and you have to work those things out again.
From the first question to the next project.
Use AI to move the work forward. Keep the design decisions yours.
Keep what you learn. Bring it into the next project.
AI helps analyse the request and context. You decide what people should do differently and whether training is the answer.
Give AI your principles, examples and project context. You decide what people need to practise and why.
AI helps check the work and find weak spots. You decide what needs to change and when it’s good enough.
Turn useful decisions and corrections into rules for the next project. You build your own design standard.
A believable wrong answer reveals more than an obvious one.
“Ignore the email completely.”
Almost nobody ignores an urgent email from their manager. This option gives the learner little to think about.
“Delete the email without reporting it.”
Someone might spot a suspicious email and delete it, but miss the need to report it. That makes the mistake believable.
A wrong answer should be believable enough that a competent but distracted person might choose it.
This is a worked example to show how I capture a correction. It isn’t a measured AI before-and-after case.
For one linear scenario with three decisions in my own process. AI writes; I make the design decisions. That’s my experience, not a promise that you’ll do it in an hour.
A free builder that helps you create, review and make the design decisions yourself.
[Name of the scenario builder]
This is a simplified version of the process I use myself.
Paste it into ChatGPT, Claude or Copilot. It takes you from course content to a practice scenario. At four points, you stop, review the work and make the design decision yourself.
Your first scenario will probably take longer than my hour. That’s fine. The four check moments are where you do most of the thinking.
After five projects with AI, what has changed about project six?
The builder lets you try this on one scenario. In the Learning Lab, you develop your own process across analysis, design, development and evaluation, with feedback from me on your work.
The Learning Lab is an 8-week online programme for experienced instructional designers who already use AI in real projects.
You'll build a way of working where you can use AI from analysis to evaluation, without giving it the design decisions that require your experience. What you learn on one project goes into the next one.
By week 8, you have your own process, rules and checks written down. You can use them on your next project and keep changing them when you learn something new.
A written approach for your next project, with clear points where AI supports you and where you decide.
Keep useful decisions and corrections, so you don’t have to make the same corrections on every project.
Review what AI produces against your own standard, rather than deciding from scratch what good enough looks like.
Add what you learn from each project to the way you work on the next one.
Weeks 1–3: work with the Lab's process.
Weeks 4–6: change it based on your own work.
Weeks 7–8: use it more independently and see what still needs to change.
Plan on 2–3 hours a week, plus [number and length] live sessions.
At the start, you do a short analysis without AI. In week 8, you do another one. Then you compare how you approached the problem and explained your decisions.
This gives you something else to look at besides the quality of the AI output: did anything change in the way you think about the problem?
Because I want to see the work you're doing and give you feedback myself.
With 40 people, I can't do that well.
During the eight weeks, you build your own AI-supported design process on a real or realistic project, in a small group of no more than 15 designers.
You already know instructional design. Now you want to make AI part of a way of working you can trust.
If you’re new to instructional design, or mainly want quick tool tips and ready-made prompts, this programme probably isn’t the right fit.
This is the first time I run the Learning Lab, so I can't show you results from previous groups yet.
What you do get is a small first group, feedback from me on your work and founding prices for the first places, starting at [€…].
The regular price will be €1,495.
The waitlist gets the full details and first pick of the 15 places on October 8, one week before the public launch.
Founding price from [€…]
Regular price €1,495
Max. 15 designers · October 8
Twenty years of designing learning. Still asking what people really need to do differently.
I spent ten years at TinQwise and ten as a freelancer, working with organisations from banks to childcare. I’ve built more than 200 courses.
The most useful lesson? Don’t start with the request for a course. Find out what people should do differently and why they aren’t doing it now. Sometimes training is the answer. Sometimes it isn’t.
My background is technical, with twenty years of practice rather than a degree in education. AI has been part of my design process since 2022. When a tool doesn’t do what I need, I tend to build one myself.
Practical questions about the Learning Lab.
You won't spend eight weeks going through lists of AI tools and prompts.
You use AI in different parts of your design process, decide where it helps and where you take over, and keep useful decisions for the next project.
I want the way you work to change after the Lab, not just the list of tools you know.
Yes. The Lab assumes you're already using AI and have seen where it helps and where it gets in the way.
We're not starting with an introduction to ChatGPT, Claude or Copilot. You'll work on how AI fits into the way you already design learning.
Plan on 2--3 hours a week, plus [number and length] live sessions.
This is the first cohort, so I'm not going to give you a made-up ROI number.
What I can tell you is what changed in my own work. One practice scenario that used to take me days now takes me about 30--60 minutes because I no longer start from scratch and repeat the same corrections.
Your result will depend on your projects and how you use the process.
Yes. The process isn't tied to one AI tool.
What matters is what you give AI, where you use it, which decisions you keep and how you review the work.
Many designers can't. You can work with public content or practice material instead. You don't need to upload confidential client data.
You can work on a real or realistic project. You don't need confidential client material to take part.
No. If you already use ChatGPT, Claude or Copilot in your work, that's enough. You won't learn to code.
Yes. You'll get an invoice and a one-pager you can share with your manager.
[Guarantee / cancellation terms if you choose to offer them.]
Try the approach on one scenario, or build your own process in the Learning Lab.
Start with the free scenario builder to explore the approach yourself. For eight weeks of practice with a small group and personal feedback from me, join the Learning Lab waitlist.
The waitlist gets the full details and first pick of the 15 places on October 8.