Consider Marcus, a software engineer in Indianapolis who has not written a straightforward line of code in years. His job, officially, is backend infrastructure. In practice, he spends his days making judgment calls: which abstraction holds up under scale, which tradeoff is acceptable, which system design will still make sense to the next team that inherits it. His output cannot be fully specified before he begins. He is evaluated as much on his thinking as his execution. By any rigorous definition, he is doing creative work. His employer has no idea how to pay him for it, and an AI vendor is currently pitching a tool that claims to do it for a fraction of the cost.
The labor market has a labeling problem. We still talk about creative jobs as if they belong to a narrow class of people, usually artists working at the cultural margins. That framing made sense in an earlier era. It no longer describes the economy most Americans actually work in, and it dangerously obscures who is most exposed as artificial intelligence moves from novelty to infrastructure.
A more useful definition begins with function rather than industry. Creative work is the act of turning ambiguity into something usable, valuable, or meaningful. By that measure, the category includes not only musicians and writers, but programmers, product managers, data analysts, and a growing class of knowledge workers whose primary task is synthesis. The boundary did not disappear overnight. It eroded gradually as technology lowered barriers and expanded reach. What remains is a market where the ability to interpret, synthesize, and create under uncertainty is no longer exceptional. It is expected, and increasingly, it is threatened.
Creative work is the act of turning ambiguity into something usable, valuable, or meaningful.
Music offers the clearest illustration of how the terrain shifted in the first wave. The barriers to producing a song collapsed. Distribution became nearly frictionless. The constraint moved from access to tools to access to attention. The modern musician became not only a creator but also a strategist, a marketer, and in many cases a small business owner managing distribution and audience relationships. The skills required began to look like the skills required to run any knowledge-based enterprise. The musician and the product manager ended up working from the same playbook, even if they never compared notes. Both now face a second disruption layered on top of the first.
Programming makes the convergence visible from the other direction. At a surface level, coding appears procedural, lines of syntax producing predictable outputs. That description misses the real work. Good engineers design systems under constraint, choose abstractions that shape how others build on their work, and reduce complexity into forms that people can actually use. The difference between a passable system and an elegant one rarely comes down to correctness alone. It comes down to taste, structure, and judgment. These are not auxiliary qualities. They are the work, and they are precisely the qualities that AI tools currently claim, however imperfectly, to approximate.
Designers, video editors, game developers, and data storytellers all fall within the expanded category. So do many roles inside organizations once considered routine. A marketing analyst interpreting customer behavior, or a policy researcher synthesizing complex data into recommendations, now performs work that is creative in both method and outcome. According to the Bureau of Labor Statistics, knowledge and information workers account for roughly forty percent of the American workforce, a share that has grown steadily for two decades. The expansion of that category is one of the genuine economic achievements of the digital era. The tragedy is the timing.
Disney’s recent announcement of roughly 1,000 additional layoffs, concentrated in its marketing division, is instructive on this point. The company has reduced its workforce by more than 8,000 positions since 2022. That the cuts now fall on marketing, the department responsible for translating creative product into audience engagement, reveals something the headline numbers do not.
The work of connecting creative output to attention is not disappearing. It is being reorganized, automated in part, and increasingly absorbed by platforms that did not exist a decade ago. The creative workers who built careers inside that studio system are not being replaced because they failed. They are being replaced because the economics shifted faster than any individual could reasonably have anticipated.
Disney is the world’s most recognizable creative enterprise, and even it is cutting the roles that bridge creation and audience. That is not evidence that creativity is being devalued. It is evidence that the institutional labor market is beginning to absorb the AI transition in the same way it absorbed earlier waves of automation: by restructuring at the top and leaving individual workers to manage the consequences on their own.
This is where sympathy and clear-eyed analysis have to coexist. Many of the workers most exposed to AI displacement are not people who ignored the warnings or failed to adapt. They are people who did exactly what the economy asked of them over the past twenty years. They learned synthesis and judgment. They built careers around interpretation and creative output. They moved into the knowledge economy because every signal told them that was where durable work lived. Those signals were not wrong at the time. The window simply closed faster than anyone with a mortgage and a decade of specialized skills could navigate.
The tools required to produce creative work have never been more accessible. A laptop and an internet connection remain sufficient to enter the field. At the same time, AI systems are increasingly capable of producing first drafts, generating visual assets, writing functional code, and summarizing complex documents at a speed and cost that compress the market for human effort in precisely those tasks. The initial expansion of the AI economy created genuine demand for creative and technical workers. The contraction that follows a technology bubble rarely distributes its damage as evenly as the growth distributed its gains.
The implications press hard against institutions that have not kept pace. Education systems still emphasize procedural knowledge and standardized evaluation. Hiring practices rely on credentials that do not fully capture creative capacity. Compensation structures struggle to account for output that is difficult to measure in advance, and retraining programs remain chronically underfunded relative to the scale of the transition underway. These systems were designed for a labor market defined by repetition and predictability. They are being asked to operate in one defined by uncertainty and originality, and the workers caught in the middle did not design the mismatch. They inherited it.
We can continue to describe the creative job market as a specialized corner of the economy, populated by people who chose an uncertain path and should have known better. Or we can acknowledge what it has quietly become: the primary mode of labor for a large and growing share of the American workforce, poorly mapped, poorly protected, and now facing a technological transition for which neither they nor their institutions were adequately prepared.
The problem is not that AI exists. The problem is that we have allowed the benefits of the AI transition to concentrate at the institutional level while the costs disperse across individual workers who have no equivalent leverage. That is a policy choice, not an economic law. Portable benefits that follow workers rather than employers, retraining investments scaled to the actual size of the disruption, and labor market data systems capable of tracking creative work as a genuine economic category rather than a rounding error would constitute a serious beginning. None of these require us to slow the technology. They require us to decide that the people the technology displaces are worth the same structural attention we gave steelworkers in the 1980s and auto workers in the 2000s.
Marcus in Indianapolis is not waiting for a think tank to name his problem. He already knows what it is. The question is whether the institutions that shaped the economy he built his career inside will move before the window closes entirely. History suggests they will move late. It does not have to suggest they will not move at all.