Shopify CEO Tobias Lütke warns that employees are dumping unvetted AI output - what he calls “slop grenades” - on their colleagues, creating more work rather than saving time. After previously setting AI use as a baseline expectation, he now reproaches workers for letting generative tools produce long, sloppy emails and convoluted code that require others to clean up. Duolingo’s founder has made a similar concession: an “AI-first” push produced scalable but low-quality content, with an estimated portion of stories becoming unusable when generated at volume, underscoring that human creativity and oversight remain essential.
Researchers have labeled this phenomenon “workslop” and quantified its costs. A survey of 962 U.S. full-time desk workers found 52.7% admitted to sending polished-looking but flawed AI work, while 38% reported receiving it and spending an average of 3.4 hours per month fixing it - up from 2 hours and 40% encountering workslop the prior year. Workslop often appears legitimate but lacks necessary components (broken links, needlessly complex code), erodes trust, and damages collaboration: recipients rate senders as less competent and 36% want to avoid future teamwork. Estimated productivity losses translate to roughly $186 per employee per month and as much as $9 million annually for a 10,000-person organization.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.