Sloptimism
A hot take on AI slop
I have a miniatures Instagram account that I carefully curate to ensure it’s solely a feed of other miniaturists. Every once in a while, I’ll see posts that make me do double takes and ask the question:
“Whoa…is this real or a mini?”
And, to my delight, I’ll discover that what looked like a full-size room while scrolling is actually a tiny scene created by a fellow maker. They successfully tricked my eye into believing I was seeing something full-scale. Reality: distorted! Mind: blown!
But there’s another question that’s entered the mix, with increasing frequency:
“Is this AI?”
Yep. AI slop has seeped into my carefully curated feed.
What’s slop?
If you’re unfamiliar with the term AI slop, think of all the images and videos of slightly mystical looking animals doing cute things, political leaders dancing on beaches, anthropomorphized cats navigating major life events, or people shapeshifting into vegetables that have been circulating on Facebook, TikTok, and other media outlets. You’ve probably seen some slop. The slop I’m seeing in my feed is usually not as egregious, but it nonetheless distorts reality in a much less delightful way than the optical illusion of a realistic miniature.
Don’t get me wrong. There are interesting use cases for AI images and videos. They reduce the barrier to content creation and can enable rapid exploration of ideas. They’re efficient and “cheap”—if you ignore the energy needed to power AI models, or the harm done to artists whose work has been scraped into training data without consent or compensation. And there are some fine artists that are doing thoughtful, creative work with the technology. So I’m not anti-AI, in general.
But a lot of what’s flooding our feeds these days can feel like tasteless gruel: lacking in depth, overwhelmingly bland, churned out fast to be algorithm-friendly, and stripped of any real human perspective. Hence the term slop.
AI slop is everywhere, and honestly, at the risk of sounding dramatic: it makes me worried about the future of humanity.
Why? To quote Ruby Justice Thelot:
“AI Slop is what emerges when the technological ease of creation meets our human appetite for rapid, uncomplicated entertainment.”
Thelot is a designer, professor, and cyber-ethnographer (a fancy way of saying “researcher of online communities and behavior”) and this week, I did a deep dive into his essays and lectures exploring culture, technology, and design. They have the esoteric phrasing and layers I associate with classic philosophy, but they’re about TikTok, being online, and, of course, AI slop. I am obsessed, so of course I had to write about it. Mostly so I can also pass along my new favorite word, which I believe is a Thelot original: sloptimist.
More on that in a few.
Slop creep
John Oliver did a somewhat depressing rant a few weeks ago about how AI slop has become a self-propagating phenomenon. In addition to blurring reality, some folks wonder: How can I, too, become rich by prompting AI to create more slop and monetizing it? Find out by taking a class from an AI cat in a hat for only $79!
Oliver also talked about Pinterest, and how it was once a delightful source of creative inspiration, but has been dealing with a big slop problem. I didn’t believe it, but a search for “garden ideas” on Pinterest pulls up an endless scroll of generated images, creating a sea of sameness despite their hyperreality, with many discreetly labeled as “AI modified.”
After watching John Oliver, my husband and I turned to each other and said: “We’re cooked.”
And I’ve had slop on the mind ever since.
The philosophy of slop
Ruby Justice Thelot shares similar concerns to John Oliver, though articulated poetically and with more philosophical references.
He argues that part of AI slop’s appeal, beyond being algorithm-friendly, is that it’s simpler and easier to consume than traditional art and imagery, which “requires dedicated time, thoughtful engagement, curation, and sustained emotional and intellectual investment.” He also notes that traditional art tends to “be divisive, frequently provoking disagreement or discomfort rather than [AI slop’s] broad, effortless appeal.” As an art lover and maker, this pained me. But as a human living in 2025, I sadly get it.
Thelot estimates that “The Flippening”—the point when there are more AI- than human-generated images—will occur in 22 years. At that point, a future generation of humans will grow up in a world where the majority of the images they see won’t be human-made, or even human-referenced. Because at that point, the images in an AI’s dataset will predominantly be AI-generated, leading to a fully recursive, ouroboros scenario where generated images aren’t referencing human-made content anymore.
But maybe it will be sooner than 22 years—I mean, you can already take a class on how to make AI slop based on AI slop. Gulp. Like I said: Cooked!
Sloptimism
Thelot also has a semi-dystopian theory that AI slop comes with a hidden benefit (his self-described “sloptimist” take): it strips power away from images in our image- and social-media-obsessed culture:
“The endless flow of trivial, instantly consumable images reduces the the individual image's power to convey lasting symbolic meaning or affect meaningful cultural transformation...the very qualities that make AI slop culturally corrosive—superficiality, low cultural half-life, speed—may also offer a radical protection…If images cannot hold our collective gaze or imprint lasting meaning, then they cannot inflict symbolic death.” - Ruby Justice Thelot
A bit heavy, eh?
As someone who’s deeply visual, the idea of images becoming less real or meaningful and becoming “powerless” gives me a lot of concern. It doesn’t exactly foster sloptimism. To quote my co-founder, Ben:
“I'm thinking about all the major social changes that were sparked by images, and how bleak the world seems if that lever for change just disappears.”
Yuppp.
But instead of simply repeating “we’re cooked,” I’m trying to take Thelot’s lead and look for a silver lining in the proliferation of slop—or at least figure out how I’ll coexist with it, because I don’t think we can stop the slop.
Perhaps I, like Thelot, can learn how to stop worrying and love the slop. And become a sloptimist.
Because maybe the sheer flood of algorithm-friendly slop will eventually push people to slow their doomscrolling and crave more reality and connection. To touch grass (or the craft mat). To make something real, with their hands. Luxuriating in tactility. Creating images and objects with raw, indisputable power. Experiencing the joy of making (which is why we’re building Practice). That future doesn’t sound so bad.
And I’ll definitely get more accustomed to asking, “Is this real, mini, or AI?” as I scroll. I’ll squint and try to tell the difference, and then feel even more delight when I discover it’s not real or AI, but mini. Something made by a human, full of depth, flavor, and imperfection.
Ohh yes the Pinterest issue has become a huge problem in my work! I use Pinterest to curate inspiration boards for client garden design projects, but there is a major ai problem there now that makes it really difficult to navigate (and the ai images feature things like plants that literally don't exist, or insane looking structures that literally couldn't be built).