Ken, a copywriter for a large, Miami-based cybersecurity firm, used to enjoy his job. But then the “workslop” piling started up.
Worksop is an unintended consequence of the AI boom. It’s what happens when employees use AI to quickly generate work that appears polished – at least superficially – but is in fact so flawed or inaccurate that it needs to be heavily corrected, cleaned upor even completely redone after it’s passed on to colleagues.
For Ken, the problem started after his company’s CEO laid off several of his colleagues and mandated that remaining workers use AI chatbots, saying it would boost their productivity. While initial drafts were a breeze to create, Ken and his co-workers had to spend more time rewriting, correcting errors and resolving disagreements between each other’s chatbots than if they had never used AI at all.
“Quality decreased significantly, time to produce a piece of content increased significantly and, most importantly, morale decreased,” said the copywriter, who spoke under a pseudonym for fear of losing his job. “Everything got a whole lot worse once they rolled out AI.” Ken said the company’s executives shifted the blame to staff when they pushed back about AI-fueled productivity decreases.
Ken’s experience reflects an emerging divide between employees and their leaders when it comes to AI: a recent survey of 5,000 white-collar US workers found that 40% of non-managers say AI saves them no time at all at work, while 92% of high-level executives say it makes them more productive.
So what’s causing this workslop deluge? The answer is more complex than being simply a case of workers cutting corners. The real driving force connects back to the C-suite.
Companies have spent billions on enterprise investment in generative AI. Some of them, like Block, Amazon, Dow, UPS, Pinterest and Target, have laid off human workers at the same time, attributing the cuts to AI’s potential productivity. Workers who remain feel pressured by their employers to use AI to produce more work, often with little guidance or training. A disconnect separates executives giddy about generative AI from workers – who are finding that AI merely makes their jobs harder.
“People are being told to use AI, often without direction or support,” said Jeff Hancock, a co-author of the study that coined the term “workslop”, and a Stanford researcher and BetterUp scientific advisor. While Hancock believes that generative AI could eventually power tools that help workers improve efficiency, in many cases, the incorporation of AI is having the opposite effect.
Hancock’s study, which is not yet peer-reviewed, surveyed 1,150 US desk workers, a subset within the total 5,000. The researchers found that 40% of workers had encountered workslop within a month, and then spent an average of 3.4 hours a month dealing with it – which the study estimates adds up to $8.1m in lost productivity for a 10,000-person organization.
Kelly Cashin, a freelance product designer, told the Guardian she encounters workslop often. “It seems to be common to just copy and paste a bot’s message directly into chats or emails,” she said. At times, when she’s confused by work a colleague sent her, they’ll respond, saying: “Yeah, I’m not sure what AI meant by that” – meaning they’re effectively outsourcing judgment to the chatbot. “Although it is personally frustrating, I understand why people do this. There’s a lot of pressure to increase productivity compounded by serious uncertainty in the job market,” Cashin said.
Philip Barrison, a University of Michigan MD-PhD student who surveyed staff while embedded in primary care clinics, found a similar workslop issue cropping up for medical staff who had been encouraged to use AI to generate email replies to patient questions. That approach was meant to save clinicians’ time.
“Based on reporting and my own observations, it doesn’t,” Barrison said. Instead, many of the workers he spoke to described a lot of editing labor, frustration and concerns about data security and patients receiving AI-assisted emails with errors. Because the AI tools are optional, “once they get past the novelty (of the AI), they start ignoring it”, Barrison said.
One reason employers are pushing generative AI in workplaces is because many companies are aiming to reduce their labor costs after investing in the tech, says Aiha Nguyen, who leads the Labor Futures program at the Data & Society non-profit research institute. But those investments haven’t paid off, or at least not yet. One of the conclusions of an often-cited MIT report found that 95% of firms aren’t seeing returns on their investments in AI. Other recent assessments from the software giant SAP and the professional services consultant and firm Deloitte report a larger fraction of businesses generating returns on investment, but they are still the minority. Businesses expect – or hope – better returns will materialize after two to four years, which is rather slow for technology investments, according to the Deloitte report.
“The problem is, generative AI is often being presented as a general-use tool that can do anything, but the reality doesn’t work that way. So what could be creating part of the workslop is (AI’s) unclear mandate or use case,” Nguyen said.
AI has become a sticking point as unionized workers negotiate the terms of new contracts, Dan Reynolds, research economist of the Communications Workers of America, said. Unions are demanding clearer mandates for the tech, and more worker input and control over how it’s used.
“Firms are pretty open about using AI to streamline operations, and so a natural response is to interrogate what those tools can actually do and the power dynamics that surround their use,” said Sarah Fox, director of Tech Solidarity Lab at Carnegie Mellon University.
Fox said she was skeptical when firms say they are deploying AI within their companies to improve productivity and efficiency and to help workers be better at their jobs. “Actually that obscures larger changes to labor dynamics,” and reduces workers’ autonomy rather than empowering them, she said.
