Facebook’s AI slop hellscape is already here

Hello! Thank you, as always, for reading Plugged In. A quick scheduling note: Starting next week, this newsletter will land in your inbox on Fridays instead of Wednesdays. See you then.

Back in December, a Meta executive catalyzed a doozy of a kerfuffle by saying the company expected AI-generated profiles to exist on its platforms “kind of in the same way that accounts do.”

“They’ll have bios and profile pictures and be able to generate and share content powered by AI on the platform. . . . That’s where we see all of this going,” VP of generative AI Connor Hayes told the Financial Times. The reaction was . . . not great (Futurism’s headline: “People Are Disgusted by Facebook’s Plan to Deploy AI-Powered ‘Users’”).

Hayes specified that Meta expected this development to happen “over time,” and though it didn’t sound the least bit appealing to me, I also thought it wasn’t an immediate threat. But in recent days, a bevy of AI-generated people have overwhelmed my Facebook feed. They are, I’m sure, a cruder manifestation of the idea than the AI members Hayes was talking up. But their presence has left me even warier of Meta willfully reimagining a social network around synthetic personalities.

The posts in question come from a variety of AI slop Facebook pages I didn’t ask to follow. The page names, such as “Nature and Animals,” “The Newstoday,” and “Dogs World,” have little or nothing to do with the topics of the posts. And those topics are, well, bizarre.

How bizarre? Something like half the posts I’ve seen involve AI-generated images of senior citizens—as old as 120—showing off birthday cakes they’ve baked themselves. Most of the others relate to talented craftspeople who have fashioned elaborate sculptures out of materials as diverse as wood, ice, and vegetables. A smattering are then-and-now shots of happy people in photos taken decades ago and in 2025. A few show distressed wild animals getting rescued by humans.

Much of the imagery is obviously synthetic; some of it is a tad more subtle. But all of it is meant to tug at the heartstrings, often in nakedly manipulative ways. A pretty high percentage of the cake bakers and craftspeople explain that their accomplishment has gone unacknowledged. Sometimes they look downright morose about the lack of love. That’s presumably meant to get Facebook members clicking, which they do—sometimes to the tune of thousands of comments and tens of thousands of likes.

Along with inserting these posts into the feeds of users who didn’t request them (like, for instance, me), Meta tries to goose engagement via its Meta AI bot. An image of a bearded gent who carved a crib for his grandchild—only to discover that “Nobody Likes It :(”—is accompanied by suggested questions such as “Why is it unliked?” and “Baby’s reaction to crib.” None of the answers the AI generates are of even the slightest value, creating a perverse feedback loop in which Meta is feeding one form of bad AI into another to generate even more slop.

Which is not to say that I didn’t find this material briefly transfixing when it first found its way into my feed. Craving more understanding of what was going on, I clicked on some of the posts to read the comments. An alarming percentage of them showered the purported posters with birthday wishes, compliments on their artistic creations, and general good vibes. Either the Facebook members who left the comments had been fooled or—worse—they were happy to be happy about the accomplishments of AI-generated characters going through slight variations on a handful of maudlin scenarios.

I did savor the delightfully cynical comments from those members who saw through the whole thing. “So what?” snapped one member in response to a post involving one particular cake-baking centenarian. “I’m 199 years old and I made my cake with peach cream and filling and I started decorating cakes when I was 6 months old.”

After a few days of this, I got worried that engaging with these posts at all had something to do with them being there—especially when the onslaught not only continued, but intensified. Maybe Facebook took me as liking them (though I never, you know, clicked the Like button) rather than merely being a victim of my own morbid curiosity. So I started using the “Not Interested” option to tell its algorithm I didn’t want to see these kinds of items. A day later, they seem to be gone, though I’m not ready to declare them eradicated.

At its best—as when group moderators take their jobs seriously—Facebook is still wonderful. It might even live up to Mark Zuckerberg’s platitudes about its mission being to connect the world. These AI posts have nothing to do with that. Stripped of its humanity and stuffed with generative AI, Facebook is the junkiest digital junk food imaginable.

And the worst part is that the bad contaminates the good. As AI-generated fake then-and-now photo comparisons began showing up, I started skimming right past similar ones posted by people I’m following. Only after slowing down and assessing them more carefully did I realize they were their real family members, not more empty calories. I felt like I’d been left to moderate my own feed—which, come to think of it, is an official Facebook policy these days.

Maybe Meta will somehow institutionalize AI profiles in a way that adds value and leaves no member confused. But I can’t help but wonder: Instead of welcoming AI-generated members onto its platforms, might the company be better off doing everything in its power to guarantee that the only people you’ll run across are, indeed, actual people?

You’ve been reading Plugged In, Fast Company’s weekly tech newsletter from me, global technology editor Harry McCracken. If a friend or colleague forwarded this edition to you—or if you’re reading it on FastCompany.com—you can check out previous issues and sign up to get it yourself every Friday morning. I love hearing from you: Ping me at hmccracken@fastcompany.com with your feedback and ideas for future newsletters. I’m also on Bluesky, Mastodon, and Threads.

More top tech stories from Fast Company

Uber will now pair Austin riders with Waymo self-driving cars
Austin riders can be matched with a Waymo all-electric Jaguar I-PACE. Read More →

Mozilla’s new message: We’re the only browser not backed by billionaires
Mozilla’s new CEO outlines coming browser features and her hopes for regulatory action. Read More →

TikTok’s ‘airport theory’ dares you to arrive just 15 minutes before your flight
A viral trend has travelers cutting it dangerously close, claiming you can breeze through security in minutes. Some make it. Others? Not so much. Read More →

Trump’s crypto reserve is a payoff for loyalists
A crypto stockpile rewards supporters while creating volatility for the government. Read More →

AI Chatbots have telltale quirks. Researchers can spot them with 97% accuracy
A new study shows that different LLMs have distinct writing styles—making AI-generated text easier to detect. Read More →

Elon Musk’s bumbling X posts are inadvertently teaching everyone how government works
Thanks to replies from journalists and experts, Musk’s X feed has become an accidental crash course in civics and a variety of topics. Read More →

https://www.fastcompany.com/91289861/facebook-ai-slop?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss