Two people recently told me they needed to "learn AI" before they could use it.

One is a friend who does community organizing. She's also a mom of a toddler, so her free time is exactly zero. She'd subscribed to an AI tool, opened it a few times, couldn't figure out how it could help with her work, and cancelled. She assumed the gap was knowledge, that she needed to learn the tool before she could use it.

The other is a neighbor who works in corporate finance. She's been hearing she needs to "uplevel on AI" for her role and was about to spend a few thousand dollars on a course. Same assumption: learn first, apply later.

Here's the thing about "learn AI" as a starting point: there's no finish line. There's a new model every month, new tool every week, new how-to guide every day. If your starting point is "understand AI," you're on a treadmill. You absorb concepts but before you can try to connect them to your actual work, things shift.

In both cases, the thing that actually clicked wasn't explaining AI. It was looking at their actual work and saying "this specific thing, right here, AI can help with that." The relevance came first. The learning followed.

There's an entire ecosystem around "learn AI, then apply it", but I think for most people, the order is backwards.