Even Microsoft Execs Are Annoyed by Terrible AI-Generated Work

A LinkedIn exec is fed up with AI slop, which might strike some as ironic given that LinkedIn itself has struggled to keep AI content in check.

In a recent post , Ryan Roslansky, a Microsoft EVP responsible for LinkedIn, Microsoft Office, Outlook, and Teams, complained that someone had sent him "what was clearly an AI-generated doc." After realizing it was AI, he actually asked his AI to summarize it, "and ended up with nothing that brought new thinking to the idea."

This, he argues, is "a doom loop" that could leave us with "a very expensive way to avoid writing and reading in an ocean of sameness, genericness."

Microsoft has pushed its Copilot AI hard, to the point that it recently pulled back on some integrations following complaints. On LinkedIn, the influx of AI tools has led to a flood of AI slop posts. It got bad enough that LinkedIn added a feature where you can click to report something that " seems like AI slop ." (And over a million people have taken advantage of that button.)

But while execs like Roslansky and Xbox CEO Asha Sharma have bemoaned "soulless AI slop," Microsoft CEO Satya Nadella is not a fan of that characterization. "We need to get beyond the arguments of slop vs sophistication," he said in March. Still, Roslansky's LinkedIn post earned the thumbs-up from fellow Microsoft employees, including President Brad Smith.

Roslansky is unlikely to ditch AI. He released a book in March, titled Open to Work, about how to make the most of AI in the workplace. (Presumably, one of the rules is to avoid maxing out your tokens .) But he's correct that replacing human communication with AI summaries is no way to run an organization. He also touches on a real concern of AI regression: who wants to read things that sound exactly the same?

PCMag and Yahoo may earn commission from links in this article.