Why don't AI courses stick?
AI courses don't stick because they deliver information once and then send people back to unchanged work. Skill needs practice and a reason to return; a course provides neither past the final slide. On top of that, the specific tips date quickly as the models improve, so even what was learned goes stale.
Learning science has said the same thing for decades: without spaced practice and application, most of what is taught is forgotten within days. A one-off AI course runs straight into that wall.
AI adds a second problem on top. Because the tools change so fast, a course teaches a version of the truth with a short shelf life. Practice on whatever tool is current avoids both traps, because the habit refreshes itself.
Move it into the flow of work: short, repeated practice on real tasks, with the gain made visible and recorded. Application and repetition are what turn a lesson into a lasting skill.