A version of this story appeared in CNN Business’ Nightcap newsletter. To get it in your inbox, sign up for free here.
All right, y’all: I’m taking a tiny kernel of my generative-AI skepticism back and giving the technology a little credit where it’s due. At long last, it has inspired something for the workforce that actually rules.
That something is not a revenue-generating tool or productivity-enhancing magic wand, sadly, but rather a fun little neologism: “workslop.”
For the uninitiated, this is the buzzword making the rounds this week after the Harvard Business Review published research from Stanford and BetterUp Labs that details an epidemic of nonsensical AI-generated work that “masquerades as productivity” and “lacks real substance.”
Workslop is drivel that looks like some sort of finished product from a white-collar job, but, in reality, it’s just gobbledygook.
Workslop, much like Shrimp Jesus or those big-eyed crying cats clogging your social feeds, has a patina of human craftsmanship. Think slick PowerPoints, official-looking reports with polysyllabic bits of jargon, lines of computer code that look like, well, usable code. But then humans who understand the actual work are left scratching their heads when the project “lacks the substance to meaningfully advance a given task.”
While some people are using AI tools to “polish good work,” others are using them to “create content that is actually unhelpful, incomplete, or missing crucial context about the project at hand,” the researchers wrote.
Naturally, that means more work for someone else to fix. Of the 1,150 US-based employees researchers surveyed across various industries, 40% report having received workslop in the last month.
One director who works in retail told the researchers: “I had to waste more time following up on the information and checking it with my own research. I then had to waste even more time setting up meetings with other supervisors to address the issue. Then I continued to waste my own time having to redo the work myself.”
This isn’t just annoying for whoever’s receiving the workslop — it actually costs companies money. Employees reported spending an average of nearly two hours dealing with each instance of workslop. Researchers calculated, based on participants’ self-reported salaries, that these incidents amount to an “invisible tax” of $186 per month. For a company with 10,000 people, they estimated workslop costs more than $9 million a year in lost productivity.
So, to recap: Not only are AI tools failing to increase revenue across the board for companies that have adopted them (as a recent MIT study found), it seems, conversely, that companies are losing money on them.
Which is not what you want to see when our entire economy is so perilously dependent on companies and investors pouring previously unfathomable amounts of money into the technology.
In the HBR report, researchers wrote that “the insidious effect” of workslop is that it “shifts the burden of the work downstream.” And that’s not wrong, but it doesn’t go nearly far enough. Having to spend your day punching up a crappy slide deck your colleague’s chatbot spit out is annoying, to be sure.
But there’s a deeper dread that comes from toiling in workslop: We’re doing so in a cultural moment where the titans of Corporate America can’t seem to stop talking about how the technology is so powerful it’s bound to replace the very people it’s been foisted upon.
AI is the future, learn to use it or else it’ll take your job, say the managers who are most removed actual day to day work of any office. Amazon CEO Andy Jassy said as much to employees this summer, echoing Anthropic CEO Dario Amodei’s message that AI will (somehow, eventually, don’t ask when) lead to a “white-collar bloodbath.”
What to say, then, to the overworked office associate making barely enough to cover their rent when they save themselves a few hours and ask ChatGPT or Claude or Gemini to write a report for them? Oh, bad job, 25-year-old with six figures of student debt. Yes, we told you that you absolutely must use AI, but we didn’t mean, like, actually use it. We meant for you to do all the work you’d do anyway but add a layer of AI fairy dust so that we can justify our subscription costs and tell shareholders we’re embracing AI but obviously you must also fact-check everything it spits out.
Workslop is the inevitable (and avoidable) result of companies blindly adopting tools that don’t work simply because a handful of Silicon Valley billionaires declared that chatbots were The Next Internet while they were at the same time building literal bunkers for the End Times.
AI companies have yet to put out a product that can fully replace human workers, but they are already laying the rhetorical groundwork to blame humans when the bots fail to make businesses more productive.
Sam Altman, OpenAI’s CEO, has avoided getting pinned down on use cases for his product, often pivoting to the idea that it’s up to us, the people, to think boldly and use AI to make the next killer app.
“Just do it,” Altman said at an industry summit in June. “When things are changing quickly, the companies that have the quickest adoption speed … win.”
Just do it.
To what end? Maybe Altman can ask ChatGPT for an answer.