Artificial Degradation: Generative AI and the Deskilling of Creative Labor

Changes in the labor market resulting from the growing availability and popularity of generative artificial intelligence tools (especially LLMs) have significantly shaped the discourse surrounding the world of work. Even the creative sector, once seemingly immune to automation, has not been spared by these transformations. Artificial intelligence is being integrated into popular tools for creative professionals. AI tools have expanded creative possibilities and accelerated previously tedious tasks, such as generating subtitles or removing backgrounds. 

Harley et al. (2025) indicate that implementing Generative AI in the workplace may reduce the time spent on a task by an average of one hour and lead to a threefold increase in productivity. However, AI-driven changes also seem to have a downside. In the public sphere, graphics, advertisements, and short clips generated with GenAI have become commonplace, prompting discussions about whether AI will replace creative workers. Demirci et al. (2024) point out that a noticeable decline in job postings for creative workers was observed on one of the prominent job platforms following the widespread adoption of ChatGPT.

In my presentation, which is based on an analysis of fifteen qualitative interviews with creative professionals, I will attempt to demonstrate how these changes are perceived by the employees themselves. By "creative workers," I mean creators and conceptual workers, (such as architects). The respondents are broken down by gender and professional experience and include representatives of the film industry, architects, and graphic designers. The interviews were conducted using a structured questionnaire containing questions about creative professionals' fears, hopes, and experiences in relation to these changes. The study is based on Labor Process Theory, drawing specifically from Harry Braverman's concept of the degradation of labor (1974), Michael Burawoy's “workshop games” (1979), and Andy Friedman's dichotomy of direct control versus responsible autonomy (1977). I also accept general framework highlighted by Judy Wajcman (2017) about political inclinations of technology. She highlights that the current development of technology is largely driven by a homogeneous group of people in Silicon Valley. This may result in significant bias that does not reflect the needs and interests of humanity as a whole. Using this methodology, I will address questions about how the capital-labor relationship shapes creative workers' processes, whether and to what extent new tools lead to deskilling, how workers cope with new challenges, and how technology shapes control over work and workers' autonomy.