Claude Lévi-Strauss distinguished the bricoleur—who works with whatever happens to be available, assembling from scraps without a predetermined plan—from the engineer, who designs from first principles with optimal materials. De Certeau appropriated this distinction and universalized it: the bricoleur is not the primitive's cognitive style but the ordinary practitioner's creative mode. The cook opening the refrigerator and making dinner from half an onion, leftover rice, and an egg nearing expiration is a bricoleur. The student assembling an essay from lecture notes, overheard conversations, and vague intuitions is a bricoleur. Bricolage is the art of making-do—judging which materials can be adapted, improvising combinations no blueprint anticipated, producing something that works from resources never designed for the purpose. In AI-assisted building, the practitioner who takes the model's heterogeneous output and assembles a coherent work is performing bricolage.
Lévi-Strauss introduced the bricoleur in The Savage Mind (1962) as a figure of mythical thought: someone whose creativity is constrained by a finite, heterogeneous collection of materials accumulated from previous projects. The engineer, by contrast, designs a project and then acquires the optimal materials to realize it. Lévi-Strauss meant this as an epistemological distinction—mythical versus scientific reasoning. De Certeau saw it as a description of the human condition. Most people, most of the time, are bricoleurs. They work with what they have, not what they wish they had. Scarcity, constraint, and the heterogeneity of available materials are not unfortunate limitations—they are the ordinary conditions of creative practice.
The bricoleur's creativity consists in recognizing potential in materials that were not designed for her purpose. The leftover rice was cooked for yesterday's meal. The jar of olives has been in the refrigerator for months. Neither was intended for tonight's dinner. Yet the cook sees that they can be combined, and her judgment about the combination—the intuition that a splash of vinegar will make the dish cohere—is creative precisely because it responds to specificity. A different bricoleur, facing the same refrigerator, would produce a different meal. The materials constrain but do not determine the outcome. The practitioner's skill, attention, and biographical disposition shape what emerges.
AI-assisted creation is bricolage because the builder works with materials she did not design: code snippets the model generated, structural suggestions drawn from the model's training data, prose inflected by the model's default rhetoric. These materials are heterogeneous—the same paragraph may contain a brilliant connection, a factual error, and a generic phrase. The builder's task is to sort, select, combine, and transform this collection into a coherent work. The model provides abundance—the full pantry where de Certeau's cook once had only scraps. But abundance does not eliminate the need for judgment. It intensifies it. The unconstrained problem (what to make from everything) is harder than the constrained problem (what to make from three ingredients), because the cook must now supply the direction that scarcity once provided.
De Certeau's appropriation of Lévi-Strauss's bricoleur appeared across his 1970s writings and reached mature formulation in The Practice of Everyday Life. He extended the concept from mythical cognition to everyday practice—cooking, walking, working, speaking—and used it to challenge the Romantic myth of creation as origination. The bricoleur does not create ex nihilo. She creates from what is there, and that mode of creation is not lesser but more representative of how most human creativity actually operates.
The bricoleur works with what is available. Not ideal materials but actual materials—the scraps, the leftovers, the tools that have outlived their original purpose.
Bricolage is creative judgment under constraint. The skill lies in seeing potential in the unpromising, in knowing which combinations will hold and which will collapse.
AI provides abundance; judgment remains scarce. The model generates unlimited material. The practitioner must judge what to do with it—the harder problem.
Different bricoleurs produce different results from identical materials. The practitioner's biography, expertise, and aesthetic sensibility determine what she makes from the model's output.