This is a common brand problem. A team wants to sound firm while the ground is moving. So it makes the promise, the channel, the campaign, and the current machine into one hard lump. When one part changes, the whole thing begins to wobble like a table with a philosophical objection to lunch.
There is a calmer arrangement. Keep the customer consequence fixed. Keep the method movable.
The promise is not the apparatus
In a recent essay, Kevin Kelly argues that a period of rapid progress may not settle uncertainty; it may create new kinds of it. His proposed response is radical adaptability, held against a small set of durable virtues. It is a useful idea, but brands need a more practical version. A customer cannot be asked to love “adaptability.” It sounds like a support ticket with a scarf.
The fixed part of a brand is the consequence it means to create for someone. A bank might stand for helping a person know where their money is going before it leaves. A tool maker might stand for making difficult work feel possible to begin. Those are not taglines. They are commitments about a person’s life.
The apparatus is different. It is the agent, retail format, feature, creator partnership, pricing test, or campaign that happens to carry the commitment this year. It deserves scrutiny, investment, and perhaps a small commemorative biscuit. It does not deserve to become the promise itself.
BUREAU SYSTEM NOTE 20: The roadmap has been granted temporary citizenship. Please do not teach it the national anthem.
Better foresight gives more options, not more certainty
This distinction matters because better tools can make a plan look less provisional than it is. A 2025 OECD and World Economic Forum report, based on a survey of 167 foresight practitioners, found that many already use AI to scan for signals, analyse data, and develop scenarios. The same report flags reliability, transparency, bias, skills, and access as live problems. The tool can widen the window. It cannot certify the weather.
Recent model research gives the same warning in miniature. Google Research’s GENSTRAT study tested nine language models across more than 36,000 strategic game matches. Models close together on an overall score still showed very different capability profiles, and some were locally volatile in similar games. These are card-game environments, not a verdict on a company. They do show why a general model ranking is a poor place to hang a customer promise.
The scale is arriving before the habits are settled. The 2026 Stanford AI Index reports AI use in at least one business function at 70% of surveyed organisations, while agent deployment was still in the single digits across nearly all functions. That is not a reason to wait outside with a thermos. It is a reason to make experiments legible as experiments.
Put a pin through the right thing
For each live brand plan, make two small labels.
Fixed point. Write the customer consequence in ordinary language. It should survive a new model, a changed channel, or a rainy quarter. If removing the current technology makes the sentence collapse, it is a method wearing a promise badge.
Floating bet. Name the current way of delivering that consequence. State what it is meant to prove and what would cause a review. A useful trigger is observable: customers abandon the task, trust falls, a better route appears, or the evidence never arrives.
A brand need not change its mind every morning. The stable promise gives people something recognisable to return to. The movable plan keeps the organisation from defending last year’s plumbing as a moral principle.
Use The Fixed-Point Card when a new capability is about to become the subject of a sentence that should really be about the customer.
Continue with A Position Is Not Yet a Memory and A Brand Should Know When to Pause.