GOALS.Reference:2.1 - The whole changes a constituent action

Consider a company promising both a usable AI-assisted advisory service and staff capability to frame and question its recommendations. During a client case, an employee’s framing of the question is part of producing the advice and part of practising the promised contribution. The two results share that action but require different evidence. Client acceptance can establish usefulness of the advice; observation of the employee’s framing under the agreed assistance can support the learning claim.

If the agent now supplies the framing entirely, the service may improve while the practice opportunity disappears. Preserve the service gain and reconsider the learning arrangement: have the employee frame a suitable next case before seeing the agent’s version, compare the reasoning and receive qualified feedback; allocate the required time. If that additional result is no longer wanted, its owner can revise the development aim. If it remains promised but no opportunity or support is available, the complete arrangement remains infeasible. Counting more delivered recommendations would leave that gap unchanged.

Conversely, a difficult first attempt can reveal that the human framing criterion was too broad for the available preparation. HCD can narrow the next learning contribution while GOALS.3 revisits which later service reliance remains supported. This return changes the whole arrangement through what happens in one constituent action. It also shows why a shared work episode need not make its service and learning results interchangeable.