The hard part of teaching AI to a group is getting people to use it after the class ends.
In August 2026, I read a discussion about research on AI and learning (Hacker News, 2026-08). Students produced better homework with AI, yet scored worse on exams. The tool had removed some of the effort through which they learned.
I see the same problem in corporate training. A manager enjoys an AI workshop, then returns to the same habits three days later.
I taught a class for executives at an education media group. Here is how I built it around their actual work, and why I ended by delivering 19 different reports for 19 people.
I'm Young. I've worked across healthcare, education, and long-term care for 20 years. Education is where I started. I've written about high school students finding a bug I missed and building an AI speaking tutor.
Who was in the room
The group asked me to teach its mid-level and senior managers. They weren't engineers. Their days involved choosing stories, writing, editing, approving work, and leading teams.
They didn't need a list of prompt tricks. They needed to see where AI fit into the work waiting for them tomorrow. So we worked with their own tasks and workflows.
Why I wrote 19 different reports
I didn't want to hand everyone the same closing comment.
For each executive, I used what they made in class, what I observed, our one-on-one conversation, and their final presentation to write a personal report. Nineteen people, nineteen reports.
On delivery day, HR told me how much detail each colleague's report contained.
Why spend the time? A generic certificate says they attended. A report about their own work gives them something to return to. It reminds them what they tried, where they got stuck, and what to do next.
AI helped me process the volume. I decided what each person needed to hear. That division of work made the reports possible.
The final presentations opened the next conversation
All 19 executives presented their work. At the end, the CEO raised three organization-wide needs: governance for data workflows, a story-selection-to-approval system that fit the group's editorial style, and a self-study tool for event planning.
Those were directions for a possible next phase of work. The class had given the group a way to see its own problems more clearly and name what it wanted to tackle next.
For me, that is the point of this kind of training. The class starts the work. Trust grows when people can use what they made and bring the next problem to the table.
What I bring to an organization
| A typical AI class | How I taught this one |
|---|---|
| General AI tips | Work drawn from the group's daily workflow |
| One closing comment for everyone | A personal report for each participant |
| The work ends with the class | Presentations surface the organization's next steps |
| An instructor visits and leaves | Someone who can help solve the next problem |
My education work also includes teaching high school students through real projects, building an AI speaking assessment system, and measuring the quality of educational AI. I started as a software teacher, and that experience still shapes how I design these classes.
Questions I get
Can non-engineering executives learn to use AI?
Yes, when the work is theirs. Every executive in this class presented something they had made.
Isn't 19 custom reports expensive to produce?
AI handled much of the volume. I handled the judgment about each person.
How do you keep the learning alive after class?
Give people feedback about their own work and a real problem to carry forward. That gives them a reason to open the tool again.
I'm Young. I've spent 20 years working across healthcare, education, and long-term care, from teaching software to building AI systems. The education media group in this story remains anonymous. If you want your team to use AI after the workshop ends, book a free conversation with me.