The tech leaders who once hailed AI use as the key to unlocking every employee’s full potential are now changing their tune.
This includes Shopify cofounder and CEO Tobias Lütke. He told employees last year that using AI is “a baseline expectation,” and they should first prove they “cannot get what they want done using AI” before asking for more resources. He’s now saying Shopify employees are producing unexamined emails and code with AI and not taking responsibility for the sloppy output.
“We call those ‘slop grenades’ that people toss at each other,” he said during an interview on The Knowledge Project podcast on Tuesday. “That’s definitely a bad thing.”
Lütke said it’s easy to let AI “go nuts,” but abusing it creates more headaches for the people receiving and reviewing the work. For instance, he suggested AI use is supposed to help synthesize points in an email rather than turning it into “a big missive” that wastes time.
“You don’t really read it, and now it has to be reviewed by your colleagues, and they are like, ‘this doesn’t look right’,” he said. “You’re just letting AI do the work for you.”
Duolingo CEO Luis Von Ahn has similarly backtracked. He announced last year the company would go “AI-first,” which meant evaluating employees on their AI usage, replacing human contractors with AI, and only increasing headcount if a team couldn’t automate the required work. But in May, he told Fast Company that he had got carried away by AI demoing well in writing, but said it ultimately doesn’t match the creativity of Duolingo’s people when it scales.
“We may need to write 1,000 different stories for people to learn a language, then you’ll find that 20% of the things were just pure slop,” he said. “Whenever we scale a lot [of] things with AI, we have to really be careful that slop doesn’t get through.”
The rise of ‘workslop’
Researchers coined the term “workslop” for the phenomenon Lütke described: polished-looking AI output that ends up dragging down productivity because it needs revision.
BetterUp Labs and Stanford’s Social Media Lab surveyed 962 American full-time desk workers this year and found over half (52.7%) reported sending workslop to colleagues and it was more common in people whose organizations encouraged AI use. Over a third (38%) reported receiving workslop and estimated it cost them 3.4 hours per month on average to revise it.
What separates workslop from low-quality work done by humans is that workslop looks legitimate on the surface while lacking the components that would make it useful. Examples of workslop mentioned by the survey respondents were well-structured emails with broken links or code that was more complicated than it needed to be.
The survey found relationships also take a hit when there’s suspicion of workslop. Employees said they viewed their colleagues who sent it in as less competent and less friendly. Of those who had received workslop, over a third (36%) also reported wanting to avoid working with those colleagues in the future.
The 3.4 hours in cleanup time is an increase from last year’s survey, when 40% reported encountering workslop and said they had to spend two hours reworking it. The number pegged to revising workslop in 2025 came out to be $186 per month for single employees and up to $9 million a year in lost productivity for an organization with 10,000 people.
