When we need a new idea, we first try to increase the number of possibilities: use different words, take a different form, combine two ideas. With AI, dozens of candidates can be created in a short time. Yet we may still feel that “they are all similar.” The candidates look different, but the range of thought itself has not changed.
One reason lies in conditions we assume from the outset cannot be moved. Suppose we want to increase use of a place where local people gather. We can propose longer opening hours, more events, or more promotion. But as long as we take for granted the condition that “people must come to this place in order to participate,” ideas remain inside that boundary. If the constraint is invisible, merely increasing ideas does not widen the search space.
For creativity, what matters is not destroying every constraint. Constraints have reasons. Constraints for safety, constraints imposed by resources, and constraints that protect the rights of others should not be removed casually. On the other hand, assumptions such as “this is how it has always been done,” “this is what people normally do,” or “this is what success looks like” may have been confused with the objective itself. We first need to distinguish necessary constraints from conventional ones.
Philosophical perspectives can be used to discover these hidden constraints. Thinking about freedom raises the question, “Does increasing the number of choices really constitute freedom?” Thinking about others lets us ask, “Is increasing the number of users the only way to enrich relationships?” Thinking about time reveals the assumption, “Is making something faster necessarily an improvement?” Philosophy here does not supply answers; it makes visible the conditions we had treated as self-evident.
Once a constraint becomes visible, it can be operated on: keep it, weaken it, remove it, reverse it, or replace it with another condition. If the condition that “everyone participates at the same time” is removed, for example, a system in which people participate little by little at different times becomes conceivable. If the condition that “users receive a finished product” is reversed, a system in which users themselves create the content may emerge. New ideas can arise not from adding elements, but from changing constraints.
This method differs from mere eccentricity. Removing any constraint does not automatically make something creative. If conditions necessary to the objective are destroyed, the idea may be new but useless. What matters is checking, “What does this constraint protect?” and “Who is affected, and how, if it is removed?” Creativity is not only freedom from constraints; it is also the ability to handle constraints consciously.
The same thing happens with human concerns. “To succeed, I must choose this path.” “Once I start something, I must continue it.” “I must not disappoint those around me.” When such assumptions are fixed, the futures we can imagine become narrow. If a philosophical perspective reveals that an assumption is not a law of nature but one value judgment or social norm, room appears to consider other choices.
There is danger, however, in AI removing constraints on its own. If it judges even responsibilities or relationships that the person values to be “mere assumptions,” AI will be making value judgments in the name of creativity. What is needed is not to decide, “This condition should be removed,” but to show, “This condition may be movable; if it is moved, these changes could occur.” The human decides whether to adopt the change.
For Sai＊aus, the meaning of holding multiple philosophical perspectives is not merely to generate a large number of candidates. A constraint that appears necessary from one perspective may appear as a revisable assumption from another. If that difference can be retained, AI can show which conditions are being held fixed before increasing the number of ideas. It supports a change in the entrance to ideation itself.
For example, in response to “I want to learn more efficiently,” AI can immediately produce ten time-management methods. But if we move the constraint that defines efficiency as “memorizing more in less time,” other objectives appear: depth of understanding, reconstruction after forgetting, connection with other knowledge, and so on. The necessary methods then change as well. New ideas arise not from ingenuity within an old objective, but from examining the conditions that had constrained the objective.
Creativity is not only the ability to produce something surprising from nothing. It is also the ability to distinguish what is fixed within an existing problem from what can be moved. The better AI becomes at mass generation, the more important it becomes for humans to consider what should be changed. Attention must shift from the volume of candidates to the constraints that determine the search space.
When new ideas do not emerge, what may be missing is not ideas but awareness of which conditions can be moved. Philosophy contains a long accumulation of thought for putting taken-for-granted assumptions into words and examining their origins and roles. If Sai＊aus can use that accumulation to make constraints visible and show the possibilities and effects of changing them, AI becomes not a device that substitutes for ideation, but a partner through which humans can expand their own space of thought.
Not answers. Understanding.