What we call art has always been a tug-of-war between makers, gatekeepers, and the people looking. But right now, the loudest gatekeeper doesn’t have a degree in art history, a curatorial statement, or a pulse. It’s a sorting system that cares about exactly one thing: engagement. Not beauty. Not difficulty. Not truth. Just engagement. And that single-minded hunger is reshaping visual culture faster than most critics are ready to admit.
I’m Zara Okonkwo, and I refuse to act like the algorithm is just some neutral delivery truck. It has taste. Lousy taste, a lot of the time, but taste all the same. And its preferences are now baked right into the work. Artists—whether they cop to it or not—are making objects designed to ping a very specific neural pathway inside a machine-learning model that’s been trained on billions of thumb-stops and double-taps. It’s weird. It’s also the water we’re all swimming in.
The Engagement Aesthetic: High Contrast, Fast Read
Spend five minutes scrolling Instagram or TikTok. You’ll see it. A visual language that’s been optimized for the feed. Colours that scream—bright, saturated things that pop against the white or black backing of a phone. Faces, usually staring straight at you with an expression that telegraphs emotion instantly. Shock. Desire. Righteous anger. Symmetrical compositions that read as a single bold shape when the image is the size of a postage stamp on a moving screen.
None of this is a conspiracy. It’s just market response. When the algorithm notices that a particular visual structure generates more dwell time, more shares, more saves, it shoves that structure in front of more eyeballs. Artists watch. They internalize. Sometimes on purpose, sometimes without even clocking it. The result is a flattening of the visual field toward something I’ve started calling the Engagement Aesthetic: high contrast, fast read, almost no ambiguity.
Quick—think of the last genuinely murky, difficult painting you saw go viral. Not a painting of a difficult subject, but a painting that refused to resolve at a glance. It barely ever happens. The algorithm has zero patience for delayed gratification. It wants the punchline now, before you even know you’re scrolling.
The Death of the Slow Image
There’s a whole tradition of art that asks you to sit with it. Rothko’s hazy rectangles. Agnes Martin’s barely-there grids. Uta Barth’s quiet photographs. These works don’t announce themselves. They gather weight over time, through sustained looking. But sustained looking is exactly what social platforms are engineered to wreck. The feed is an endless scroll of interruptions. Each post fights for a fraction of a second before the thumb moves on.
What wins in this arena is the image that can say everything it has in that fraction of a second. Certain genres get a boost: hyperrealistic drawing that scans as a party trick, optical illusions, before-and-after transformations, art that moonlights as a political slogan. Subtlety gets punished. Abstraction gets punished. Anything that needs context outside the frame gets punished.
I’ve talked to young painters who feel trapped. They want to make slow, quiet work. But they also want to eat. And the path to commissions, gallery eyeballs, and sales now runs straight through Instagram. So they make the fast work for the feed and the slow work for themselves, praying the two never meet. But they do. The fast work becomes the public face. The slow work becomes a hobby. Over time, the distinction dissolves.
The Algorithm as Curator: Invisible and Unaccountable
Museums and galleries have always been biased. They’ve favoured certain artists, certain styles, certain identities. But at least those biases were visible and, in theory, open to argument. You could write a scathing review of a curator’s choices. You could protest a museum’s exclusion of women or artists of colour. You can’t argue with an algorithm. It doesn’t have a phone number. It doesn’t publish its criteria. It just… sorts.
Because the sorting is invisible, it naturalizes its own preferences. When a particular flavour of art keeps showing up in your feed, it starts to feel like the only flavour of art. The algorithm doesn’t just reflect taste—it manufactures consensus. A 2022 study from the Pew Research Center clocked how algorithmic feeds shape not just what users see but what they believe is popular, creating a feedback loop that can amplify certain aesthetics while erasing others entirely.
This is how we end up in a world where a specific strain of digital surrealism—glossy, vaguely dystopian, heavy on neon and chrome—becomes the default visual language of a generation. Not because it’s the most interesting or the most challenging, but because it performs well on screens. The algorithm loves shiny things. It’s basically a magpie with a server farm.
When the Platform Becomes the Patron
Historically, art was shaped by whoever was paying the bills. The church wanted piety. The aristocracy wanted flattery. The market wanted commodities. Now, the platform is the patron, and it wants content—a steady, unending stream of it. The platform couldn’t care less if the work is good. It cares if the work is frequent. Artists who post daily get rewarded with visibility. Artists who post when the thing is actually finished get punished with silence.
This has given rise to a whole new mode of art-making that’s essentially performance. The process becomes the product. Sketchbook tours, studio vlogs, time-lapse videos of paintings being made. The finished piece is almost an afterthought. What matters is the content pipeline. I’ve seen artists with massive followings whose actual work is mediocre at best, but their content about the work is compelling. The algorithm has made them famous for being artists, not for their art.
And look, I’m not a purist. I don’t believe there’s some sacred boundary between making and showing. Artists have always documented their process. But when the documentation becomes the main event, something shifts. The work starts being made for the documentation. Angles are chosen for the camera, not for the person standing in the room. Colours are selected for how they photograph on a phone sensor. Scale is determined by how it’ll look cropped to a square.
The Platform-Specific Vernacular
Different platforms breed different aesthetics. TikTok’s algorithm favours video that hooks you in the first half-second, which has produced a whole genre of paintings that start with a dramatic slash of colour or a face emerging from chaos. The payoff has to be immediate and visceral. On Instagram, the carousel post has become an art form in itself—images sequenced to tell a story, or to hide the best piece at the end so people swipe through. The swipe is a metric. Artists are now composing in swipes.
Pinterest, which is technically a discovery engine more than a social network, has its own gravitational pull. It favours the aspirational and the instructional—art that can be used, pinned to a mood board, replicated in a DIY project. This nudges certain aesthetics toward the decor-friendly and far away from the confrontational. You don’t see much Francis Bacon on Pinterest. You see an avalanche of watercolour botanicals.
None of this is exactly new. Artists have always responded to the constraints of their medium. Fresco painters worked fast because plaster dries. Sculptors chose stone based on what was available locally. But the algorithmic medium is different because its constraints are hidden and constantly shifting. The algorithm changes, and suddenly a visual strategy that worked for years stops working. Artists are left guessing why.
The Homogenization Problem
Here’s what keeps me up at night: the algorithm doesn’t just shape individual works. It shapes the entire ecosystem. When artists around the world are all optimizing for the same engagement signals, the result is a global convergence of style. Scroll through the art tags on any platform and you’ll see it—a sameness that crosses borders. The same colour palettes. The same compositions. The same emotional registers.
This isn’t about individual talent. The artists are skilled. They’re grinding. But they’re all solving the same optimization problem, and optimization problems tend to have a limited set of solutions. Real originality—the kind that feels jarring, wrong, uncomfortable—doesn’t optimize well. It confuses the algorithm. It confuses the audience trained by the algorithm. So it sinks.
I think about this every time I walk into a gallery show that looks like an Instagram feed printed and framed. The work is competent, polished, entirely without friction. It’s art that has been pre-approved by a machine-learning model. And the market loves it, because the market is just people who’ve been scrolling the same feeds as everyone else.
Resistance Is Possible, But It’s Work
I don’t want to end on a note of total despair. Artists are clever. They find ways to game the system, subvert it, or step outside it entirely. Some are using the algorithm’s preferences against it—making work that looks like engagement bait but then delivers something genuinely unsettling once you stop to look. Others are building audiences through slower channels: newsletters, physical zines, real-world communities that don’t depend on a feed.
There’s also a growing hunger for difficulty. I see it in the comments on posts that break the formula—people saying thank you for making something I had to think about. The algorithm may not reward that comment with a share, but it signals something real. A fatigue with the easy. A craving for friction.
Critics have a role here too, though I say this with some embarrassment because criticism is also getting flattened by the same forces. Long-form writing doesn’t perform well on social media. Arguments that refuse oversimplification get reduced to hot takes. But there’s still space for the kind of attention that art demands, and I intend to keep carving out that space, even if the algorithm would rather I make a 60-second video instead.
The Long View
Art has survived worse patrons than the algorithm. It survived the church. It survived the academy. It survived the market. It will survive this. But survival isn’t the same as flourishing. The question isn’t whether art will continue to be made—it will. The question is what kinds of art will be visible, and what kinds will be lost to the feed, and what that loss will mean for the rest of us who need art to be more than just content.
I don’t have a neat conclusion. The algorithm hates neat conclusions anyway. It prefers a cliffhanger, a question that keeps you scrolling. So here’s one: the next time you see a piece of art on your phone, ask yourself whether you’re liking it because it moved you, or because it was designed to be liked. The difference matters.
Frequently Asked Questions
Do social media algorithms actually change what kind of art gets made?
Yes, and in measurable ways. Artists report adjusting their colour palettes, compositions, and even subject matter based on what performs well on platforms like Instagram and TikTok. The economic incentives are real—visibility leads to commissions, sales, and gallery interest. Over time, these individual adjustments accumulate into broad stylistic trends that prioritize instant readability and emotional punch over ambiguity or slowness.
Can an artist succeed without playing the algorithm game?
It’s possible but increasingly difficult. Some artists build audiences through slower channels like newsletters, physical events, or word-of-mouth networks. Others achieve enough institutional recognition—through gallery representation or grants—that they don’t depend on social media for income. But for emerging artists without those advantages, the algorithm is often the only scalable way to get work in front of an audience.
Is the algorithm’s influence really different from past gatekeepers like galleries or critics?
Yes, in two important ways. First, the algorithm is invisible and unaccountable—you can’t argue with it or understand its criteria. Second, it operates at a speed and scale that no human institution ever matched. A curator might shape taste over years; an algorithm can reshape the visual landscape in weeks. The combination of opacity and velocity makes it a uniquely powerful force.
What can viewers do to resist algorithmic taste-making?
Seek out art through non-algorithmic channels: visit galleries and museums in person, subscribe to artist newsletters, buy physical publications, follow recommendations from humans you trust rather than feeds. When you do use social media, engage deliberately with work that challenges you, even if the platform doesn’t reward it. The algorithm responds to behaviour—if enough people change their behaviour, the signals shift.