What Is 'Workslop,' The AI Trend Workers Hate?
If you take a July report pubished by Microsoft (which has invested billions into OpenAI) at face value, you might have reason to fear artificial intelligence’s impact on your job. After analysing hundreds of thousands of conversations with its Co-Pilot chatbot, Sky News reports, the company “concluded [AI] could complete at least 90% of the work carried out by historians and coders, 80% of salespeople and journalists, and 75% of DJs and data scientists”. But if you ask the researchers behind a study published by the Harvard Business Review, generative AI tools are actually “destroying” workplace productivity. That, they say, is because it creates “workslop” – a trend that over half of workers surveyed say annoys them. What is “workslop”? The inventors of the term say it refers to “AI-generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task”. The use of AI to improve already-good work is not the problem, the publication says. “Workslop” happens when artificial intelligence is used to “create content that is actually unhelpful, incomplete, or missing crucial context about the project at hand”. It can appear in emails, reports, slides, and even code as more and more workers reach for AI tools to churn out polished-sounding tasks faster. And of the workers they surveyed, 40% say they’ve received “workslop” in the past month. One of the most “insidious” side-effects of “workslop” is that it can sometimes take a long time for the person taking it to correct or interpret the work, moving the burden of the work from the sender to the receiver. In this research, the flow runs from higher-ups to their teams 16% of the time. No wonder 53% of workers say receiving “workslop” left them annoyed, while 38% say it confuses them, and 22% say it flat-out offends them. And it’s bad news for the “worksloppers,” too: those who send it to their coworkers are seen as less intelligent by them 37% of the time, and less trustworthy 42% of the time. Almost a third of people who received “workslop” from a colleague said they were less keen to work with the sender again in the future. How much might “workslop” cost? According to the Harvard Business Review, employees spend just under two hours dealing with every instance of “workslop”. That’s not just annoying – it’s expensive. The researchers say, “Based on participants’ estimates of time spent, as well as on their self-reported salary, we find that these workslop incidents carry an invisible tax of $186 per month. “For an organisation of 10,000 workers, given the estimated prevalence of workslop (41%), this yields over $9 million per year in lost productivity.” It follows an MIT report published earlier this year, which found that 95% of companies that adopt generative AI solutions see precisely zero measurable return on their investment. “The core barrier to scaling is not infrastructure, regulation, or talent. It is learning,” the MIT report reads. “Most GenAI systems do not retain feedback, adapt to context, or improve over time.”
