Design was never the bottleneck
Jonathan Stephens
April 29, 2026
> The more I look at the state of tech, the more I recognize a specific structure operating in two distinct ways: the way business absorbs design's attempts at strategic influence, and the way AI tools absorb designers' labor while promising to liberate it.
> Design's position within business follows a logic that extends well beyond design. The work that designers do (labor), understood as a political-economic relation rather than a finite task list, will fill whatever vessel it is placed inside.
> Focused, uninterrupted work, the kind required for complex problem-solving and creative production, and the actual work that brings value to a company, declines.
> What makes this legible as a systemic condition rather than a professional grievance is the way it connects to the broader logic of contemporary capitalism. Ajay Singh Chaudhary's analysis of what he calls the extractive circuit describes a global system in which, at every node, inputs are extracted and exhausted: ecological, political, social, individual. Value is produced through speed, the shifting of risk to whatever actor is structurally least able to absorb it, and the transformation of every moment of life into something productive and profitable.
> Every serious study of AI's effects on work arrives at the same impasse. The empirical findings point clearly toward intensification, expanded scope without expanded autonomy, cognitive overload, erosion of focused work, and blurred boundaries between work and rest. But the recommended responses never follow the findings to their conclusion.
> Design's actual disciplinary power is the capacity to imagine something that does not yet exist and to make it concrete. It's a change mechanism. It is a way of intervening in the world by proposing that the world could be organized differently. That capacity has been redirected toward making existing products marginally more usable, or translating business requirements into interface decisions, or proving and re-proving value to organizations that have already decided what design is worth in the first place.
> Design skipped this step. It went directly from craft to corporate employment without building the intermediary structures that would have given the discipline some form of independent standing. Design lacked the institutional infrastructure that traditionally anchors professional autonomy, so corporate-facing frameworks like maturity models tried to earn legitimacy within employers’ terms instead of establishing a separate disciplinary frame. In that sense, they addressed recognition, not sovereignty. They were treating the symptom.
> Cooperative studios where designers own their own conditions of work. Research collectives that produce knowledge about design's material conditions rather than business cases for design's existence. Educational spaces that prepare practitioners to build alternative organizational forms rather than to compete for positions within existing ones. Professional bodies oriented not toward networking and career development but toward establishing shared standards of practice that practitioners themselves define and maintain.
> None of this can happen at the individual level. What makes change is collective action: practitioners identifying shared conditions, developing shared analysis of those conditions, and building shared institutions that embody a different set of terms.
Discussion in the ATmosphere