Exercise: survey the familiar answers

An attractor survey begins by making predictable answers visible. Choose a narrow topic and ask for several common ways of framing it. The result is a candidate inventory of conventions, not a measured frequency distribution of a training corpus. The model may omit important views or mistake a familiar-sounding formulation for a common one.

For a library example, possible framings include access to information, community space, digital assistance, collection management, and public expenditure. Ask what each framing treats as the main problem and what it tends to leave in the background. You now have a more explicit object to compare with real documents and experience.

A prompt to adapt

List six familiar framings of this question. For each, name its main assumption, one useful insight, and one likely omission. Mark any claim about prevalence as an estimate rather than a measurement.

After receiving the list, add a framing yourself. Then ask whether the new framing changes the proposed actions or only the language used to describe them. This interrupts the tendency to let the model define both the field and the evaluation of its own field.

Choose one attractor for further work. A familiar answer may be correct and useful; the purpose is to inspect its conditions, not to reward obscurity. If you want to move away from it, specify what assumption should change and what practical constraint must remain.

Carry that pair into the boundary search. Keep the original survey so that later claims of novelty can be compared with an actual starting point. The record will matter when you perform integration.