Make something the metric cannot imagine

A measure can tell you which proposal performs best against a question you already know how to ask. It offers less help when the valuable proposal would change the question. This is a familiar difficulty in creative work. A museum may know how to count visitors before it knows how to recognize a new relationship with them. A publication may know what people finish reading before it understands what has helped them begin thinking differently.

AI can make the available proposals more varied, polished, and accessible. It also makes an old leadership responsibility harder to evade: deciding what deserves to exist. The ability to generate another version leaves the purpose of the work unsettled. Some creative acts introduce a form of participation that was absent before. Once it exists, people can want it, care for it, and judge it. Its arrival enlarges what there is to evaluate.

A study by Anil Doshi and Oliver Hauser helps sharpen one part of the problem. In their short-fiction experiment, access to AI ideas improved evaluations of individual stories while making assisted stories more similar to each other. Both findings matter. Assistance can help a person create, while a collection becomes less varied. The experiment concerns a particular writing task. It establishes no general rule that AI destroys imagination or that human work escapes repetition.

The broader question for a leader is how an organization makes room for a good its present measures cannot yet recognize. Consider a hypothetical museum planning a new exhibition. The team uses AI to help prepare several tour concepts, then ranks them by expected attendance and visitor satisfaction. These could be useful comparisons. Yet every proposal assumes the same arrangement: the institution speaks, and the visitor receives a more or less satisfying experience.

An educator proposes something different. Visitors could bring an everyday object whose history unsettles one of the exhibition's assumptions. With permission, a few of those histories could enter a changing section of the exhibition. The proposal creates a new relationship: the visitor becomes someone to whom the institution must listen. This may improve the existing scores or reduce them. Its distinctive promise will remain hard to see if the team can discuss it only as another tour.

For one project this week, write a page headed A Good We Have No Place For. Begin with an experience you want to make possible for another person. Describe a scene specific enough to imagine: someone sees their account taken seriously and finds that it changes a public interpretation. Avoid beginning with the score you hope to improve. You are trying to give a possibility enough form that somebody else can respond to it.

Make a small version of that scene. The museum could invite a few willing participants to examine one object and one account before committing to an exhibition. A team designing a service could enact an encounter rather than produce another presentation about it. The artifact might be a short script, a table with two chairs, or a rough page someone can actually use. Choose a form that exposes your idea to another person's experience.

Afterward, write down what happened that your original score would have missed. Ask the participants what they found meaningful, confusing, or unwelcome. They may reject the premise. That response belongs in the record. If a new good becomes visible, describe it in ordinary language before inventing a scale for it. The museum might learn that being allowed to correct the institution matters more to some participants than supplying an additional story. That would require changing the design again.

Protect the ordinary obligations while you do this. Originality does not excuse unreliable service or give an institution permission to use people as material for its imagination. Keep the experiment small, explain its provisional nature, and let people decline. The exercise is also not evidence of a new market or a universal theory of creativity. It tests whether a particular possibility deserves further attention.

Notice whether your next decision includes a reason that the first ranking could not express. Perhaps the team changes its question from which tour people prefer to what kind of relationship an exhibition should make possible. Measurement can follow that discovery and help care for it. Leadership contributes by keeping the first question open long enough for something worth measuring to arrive. The work becomes creative when it gives the future more to contain than better versions of what you already know.