What to ignore is a strategy decision

Any competent team can generate a list of places AI might help. Give a smart group a week and they'll hand you twenty opportunities, each plausible, each with a story. The list feels like progress. It usually isn't. A list of everything you could do is not a strategy — it's a menu, and menus don't decide anything.

The strategy is in what you leave off. A prioritized list is only real if something was explicitly excluded, and the exclusions are where the hard thinking lives. Naming what the company will not pursue this year — and why — is often the single most valuable output of a clarity engagement, and the one people most want to skip.

It's uncomfortable for a specific reason: every item you cut is plausible. Nobody proposes obviously bad AI ideas. You're not choosing between good and bad; you're choosing between good and good, with a finite amount of leadership attention, engineering time, and organizational patience to spend. Twelve pilots pursued halfway return less than two pursued to production. But twelve feels safer, because no single champion had to be told no.

The operator's version is blunt: attention is the scarce resource, not ideas. A leadership team can hold maybe two or three real initiatives at once before everything slows and nothing ships. So the work isn't finding opportunities. It's ranking them honestly and drawing a line — this year, these; not these, and here's the reason. The reason matters as much as the ranking, because next quarter someone will re-propose a cut item, and “we decided not to, and here's why” is the only thing that holds the line.

When we run a clarity engagement, the deliverable clients underestimate is the stop list. The opportunity list makes everyone feel busy. The stop list is what makes the next twelve months actually work.

If your AI plan doesn't name what you're not doing, you don't have a plan. You have a wish list with better formatting.