In 2022, when AI exploded into the mainstream, I was running operations for an enterprise software company.

I was overcommitted and understaffed, leading multiple company-wide initiatives, turning around one drowning function and standing up several new functions, all at the same time.

I didn't have time for the world to change.

But though I didn't know yet what AI would mean — and literally no one had a map — I recognized this was an inflection point. So, I piled on one more commitment, and created and led our AI initiative anyway.

Five months later, we were using AI internally and we delivered a slate of new AI product features — built from the ground up.

Here's what I learned: Asking non-technical people to learn technical skills like prompt engineering was a failure.

(In fact, this is one of 27 AI failure modes I've identified.)

What did work? Treating AI like I treated the rest of my team — using the same management frameworks, the same delegation principles, the same leadership approach.

Then I taught other leaders to do the same. And I watched the doors open for them too.