Our social media feeds are already dominated by eerie, poorly made “AI slop”.
Now, the lazy use of AI tools to produce passable but unhelpful work has emerged in the form of “workslop”, a new report has found.
In an article for the Harvard Business Review, a team of researchers coined the new term for “AI-generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task”.
The report found that “workslop” is now rife across a range of industries, but particularly common in tech, and is having a major productivity and personal impact on workers.
Researchers found the accessibility and availability of AI tools today means they can be used in the workplace to produce “polished output” such as well-structured reports and seemingly articulate summaries of academic papers by non-experts.
“But while some employees are using this ability to polish good work, others use it to create content that is actually unhelpful, incomplete or missing crucial context about the project at hand,” the report said.
“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.
“In other words, it transfers the effort from creator to receiver.”
The researchers present this as a potential reason why many of the companies that have fully embraced AI are yet to see any real economic benefits of it.
A recent report by MIT Media Lab found that 95 per cent of companies have seen no measurable return on their investment in these technologies.
“A confusing contradiction is unfolding in companies embracing generative AI tools,” the study said.
“While workers are largely following mandates to embrace the technology, few are seeing it create real value.”
Productivity and social issues
An ongoing survey of 1,150 US-based full-time employees across a range of industries found that 40 per cent had received workslop in the last month.
Most of the time this poor AI content is being sent between peers, but just under 20 per cent of the time it is being passed on to managers.
“Workslop uniquely uses machines to offload cognitive work to another human being,” the report said.
“When coworkers receive workslop, they are often required to take on the burden of decoding the content, inferring missed or false context.
“A cascade of effortful and complex decision-making processes may follow, including rework and uncomfortable exchanges with colleagues.”
Respondents to the survey said that is taking on average just under two hours to deal with workslop and properly decode it.
Based on their self-reported salaries, workslop incidents are costing about $186 per month, and for an organisation with 10,000 employees, it would be slugging them $9 million per year in lost productivity.
‘Workslop’ producers seen as less intelligent
On top of the productivity impact, the prevalence of workslop is also impacting relationships between co-workers, the report found, with more than half of respondents saying they feel annoyed when receiving this type of work, with more than 20 per cent being offended.
Half of those surveyed said they view colleagues who have sent workslop as less creative, capable and reliable than previously, 37 per cent as less intelligent and 42 per cent as less trustworthy.
“Over time, this interpersonal workslop tax threatens to erode critical elements of collaboration that are essential for successful workplace AI adoption efforts and change management,” the study said.
To combat the rise of workslop, the researchers recommended being precise and careful about where AI is implemented across the workforce, and to make sure that collaboration is focused on above all else.
The use of AI in the workplace is also having a wider impact, with a study last year finding that workers were feeling depleted and uninspired, in part due to their boss’s focus on AI and automation.
Some tech firms have also been forced to backtrack after going all-in on AI, realising they still need the human touch and re-hiring some laid off staff.
