Make Yourself With AI

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI documents the make-yourself-with-AI phenomenon in its most dramatic early manifestations: Alex Finn building revenue-generating software products alone across twenty-six hundred hours without writing a line of code by hand; the Napster Station sprint in which industrial design, optics, audio routing, conversational AI, and software integration were combined by a single directing intelligence with AI assistance in thirty days; the engineers at Trivandrum each discovering in a week that they could perform work previously requiring combined team output. Each episode is the third option in action. The specification-to-implementation transaction that previously consumed weeks of meetings and document review was replaced by conversation with an AI system that could hold the full context of the project. The coordination costs between disciplines were eliminated when a single director could maintain coherence across domains with AI assistance.

The cycle frames this as the collapse of what it calls the imagination-to-artifact ratio. When the cost of converting an idea into a working product approaches zero, the entire structure of economic organization that existed to manage the conversion process—the project manager, the technical lead, the QA department, the deployment pipeline—is called into question. Each existed because the conversion was expensive, complex, and required specialized coordination. When the conversion can be accomplished by a single person describing what they want in natural language, the coordination costs that justified each of those roles are no longer incurred.

The Coasian frame does not celebrate this or lament it; it analyzes it. Every organization must now ask freshly at each cycle: which activities do we still perform through internal coordination, which through market procurement, and which can our individuals now produce themselves? The answer will be different from last quarter, and from last year, and will be different again next quarter. The make-or-buy decision has always been dynamic; the third option has made it more dynamic still.

Origin

The concept of make-yourself-with-AI emerges from applying Coase's make-or-buy framework to the actual decision-making behavior of practitioners documented in the AI literature of 2025-2026. Coase's original framework was binary because individual production capacity was bounded by the skills of a single person; the solo contractor could produce code or design but not a complete product requiring both. AI changed the production function: the AI-augmented individual now has access to capabilities that previously required a division of labor.

The most important structural feature of the third option is that it does not eliminate transaction costs so much as relocate and transform them. The costs of negotiating with a vendor disappear; the costs of directing an AI system and verifying its output emerge. The costs of internal coordination between specialists disappear; the costs of the individual's judgment in integrating across domains emerge. These are real costs, and they create a new frontier in the make-or-buy analysis: the boundary between what can be make-yourself-with-AI and what still requires specialized human judgment is itself moving, and tracking it requires exactly the empirical discipline Coase always demanded.

The concept connects to the broader restructuring Coase's framework predicts: as the third option expands, firms whose value resided in the production layer face the arithmetic that every organization now confronts. Firms whose value resided in the ecosystem layer—the data, integrations, customer relationships, institutional trust—retain their economic rationale. The production layer is migrating to individuals; the ecosystem layer is what the firm of the future will be built around.

Key Ideas

The Structure of the Third Option. Make-yourself-with-AI eliminates both the coordination costs of the firm and the transaction costs of the market, replacing them with the costs of directing an AI system: the compute cost, the prompt engineering cost, the verification cost, and the cost of the individual's judgment in directing a tool whose outputs may be excellent or subtly wrong. The net reduction in total cost is large in the domains where AI performs reliably; the cost structure is unfavorable in the domains where reliable AI direction requires the very expertise the individual is being asked to develop independently.

The Expanding Range. The boundary of what the third option can accomplish is not fixed. It expands as AI capabilities improve, as the accumulated experience of millions of users produces prompting strategies and workflow patterns that extract higher-quality output, and as the feedback loops between AI-augmented practitioners and AI development teams narrow the gaps that individual production most encounters. An activity that required specialized judgment last year may be well within the third option's range this year. The make-or-buy analysis must be rerun continuously.

The Judgment Premium. The third option amplifies existing judgment without replacing it. The experienced engineer directing AI-assisted work across multiple domains produces higher-quality output than the less experienced engineer using the same tool, because the judgment layer she provides is thicker. The third option does not democratize production equally; it amplifies the judgment that practitioners already possess. This has implications for the judgment economy: the premium on human judgment may increase rather than decrease as the third option expands, because judgment is now the binding constraint where previously it was the bottleneck alongside production.

What the Third Option Cannot Do. The third option operates on the production layer; it does not provide what markets cannot provide—trust built over time, the transmission of tacit knowledge through mentoring relationships, the maintenance of professional standards through peer community, the sense of belonging that sustains difficult work. The individual who make-themselves-with-AI gets the production; they do not get the social infrastructure. This is why the guild and similar organizational forms emerge: to provide the social layer that the third option's economics make possible but not automatic.