I have sat in rooms where the loudest person called AI a fad, a toy, a distraction from “real work.” Six months later that same person wants a private briefing because a competitor just took three accounts. Dismissing AI until you start losing clients is not skepticism. It is a delayed purchase order with interest.
Eating your words is available to everyone. The market does not care that you were witty on the way down. It cares whether you still have customers when you finally get curious.
What does “dismissing AI” look like in practice?
Not a thoughtful “show me production proof.” Dismissal is a posture: refusing to learn the difference between a chat tab and an operating layer (AI is not ChatGPT), mocking early movers as reckless, and outsourcing your judgment to advisors who still sell the course that already moved.
The tells are boring: no architecture owner, no pilot with a review date, no one allowed to be wrong in public. Meanwhile the shop across town answers every call, follows up every lead, and shows up in answer engines you still think are “just Google.”
When the bill shows up
Clients rarely send a letter that says “we left because you dismissed AI.” They say price. They say speed. They say the other guy “just feels more on it.” Underneath: response time, estimate quality, follow-up discipline, a website that actually answers questions — or a retainer that never shipped the work.
By the time you feel it in revenue, the inflection already happened. That is the same math as three hands in a room of hundreds and the contractor across town with a robot. Waiting feels neutral. It is a bet.
Failure modes of late conversion
- Panic buying. You purchase five tools in a week and call it a strategy. See bolt-on vs. built.
- Stage shopping. You hire the cheapest “AI speaker” to calm the board — which can do real damage if they are not operators ( don’t take the mic).
- Rewriting history. You pretend you were “cautious, not opposed.” Your team remembers. Trust compounds slower than tools.
Proof — and grace
Being early looks crazy until it doesn’t — I wrote that loneliness. Being late and honest still beats being late and loud. Grace under the noise is Suck It With a Smile. What I will not do is pretend dismissal was free. The Prediction Ledger exists so calls stay dated — including the ones people wish they could unsay.
What honesty looks like after dismissal
You do not get to skip the apology to the people who watched you mock the early movers. You also do not get to confuse apology with strategy. Strategy is a pilot with a owner, a budget, and a review date — then a second pilot that compounds on the first. Strategy is not a offsite where everyone agrees AI matters now that the competitor's trucks look busier.
If your advisors still coach delay, replace the advisors. The expired course has a body count measured in lost accounts. Eating your words is available. Pretending you never said them is how you lose the room twice.
One more tell: the owner who says "we'll revisit AI next fiscal year" while their best estimator already uses a private ChatGPT account to write proposals. The company is not waiting. The org chart is. When the unofficial tools outrun the official posture, you are already eating your words — you just have not said them in the all-hands yet.
Action
If you were the dismisser: say it out loud to your team this week. Then pick one workflow with clear unit economics and put a review date on a real pilot — not a tool tour. If you need the inflection argument on stage, book Jennifer or start with AI for the Trades. If you want the long-form charge to stop waiting, read Hands Up. Score public foresight at /prediction-ledger.