Researchers from consulting firm BetterUp Labs, in collaboration with the Stanford Social Media Lab, have introduced the term "workslop" to characterize low-quality, AI-generated content. This definition, outlined in an article published in the Harvard Business Review this week, refers to "AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task."
The concept of workslop is posited as a potential explanation for reports indicating that up to 95% of organizations attempting AI integration have not yet observed a return on their investment. According to the researchers, such AI-generated content can be "unhelpful, incomplete, or missing crucial context," thereby increasing the workload for subsequent stages of a task. They noted that the "insidious effect of workslop is that it shifts the burden of the work downstream, requiring the receiver to interpret, correct, or redo the work." This dynamic carries implications for industrial sectors, where AI is increasingly employed in areas such as predictive maintenance, supply chain optimization, and automated quality control, potentially affecting operational efficiency if outputs require significant human intervention.
An ongoing survey conducted by the research teams involved 1,150 full-time, U.S.-based employees. The survey indicated that 40% of respondents reported receiving workslop within the past month. This prevalence suggests a broader challenge in maximizing AI utility across various business functions, including those within manufacturing and logistics environments where timely and accurate data is critical for decision-making and automated processes.
The findings from BetterUp Labs and the Stanford Social Media Lab highlight the need for careful consideration in the deployment and management of AI tools within industrial settings. The researchers presented approaches aimed at fostering intentional AI use and establishing clear usage norms within teams to mitigate the generation and impact of workslop.
