Ghost Code: The Artists Dragging Hidden Digital Systems Into the Light

There's a version of the internet you interact with every day—the buttons, the feeds, the thumbnails, the scroll. And then there's the version underneath that one. The one you never see. It's making choices about what you read, what you buy, who you meet, what you believe is true. Most of us just call it "the algorithm," shrug, and keep scrolling.

A loose, scrappy, and increasingly urgent community of artists, coders, and activists has decided that's not good enough. They're building work that forces those hidden systems out into the open—not just to expose them, but to make you feel what it's like to be processed by one.

The Infrastructure You Live Inside

When we talk about digital infrastructure, most people picture servers in a warehouse somewhere. But the real infrastructure shaping daily life online is behavioral. It's the decision tree that decides your credit score is too low. It's the content moderation model that pulls down a protest video but leaves up something worse. It's the ad-targeting system that figures out you're pregnant before you've told anyone.

These systems aren't neutral. They encode choices—about who matters, what counts as risky, whose speech gets amplified. The problem is they're designed to be invisible. Opacity isn't a bug; it's a feature that protects the companies running them from accountability.

That's exactly the gap artists have started moving into.

Making the Machine Confess

Take the work coming out of the data visualization community over the last several years. Artists like Tega Brain and Sam Lavigne have built projects that don't just show data—they expose the assumptions baked into the systems that generate it. Their project Smell Dating, on its surface a goofy sensory matchmaking experiment, quietly interrogates how platforms reduce human compatibility to quantifiable metrics. It's funny until it isn't.

Or look at what's happening in the browser art space, where developers are building tools that make algorithmic decision-making visible in real time. One approach that's gained traction: browser extensions that visually annotate your social media feed with the probable reasons each post surfaced. Why are you seeing this? Because you watched a video three weeks ago. Because someone in your network engaged with a similar account. Because the platform is trying to keep you in a specific emotional state that increases ad engagement.

Seeing it spelled out like that lands differently than reading about it in a think piece.

Code as Material, Not Just Tool

What's interesting about a lot of this work is that the artists aren't just describing these systems—they're using code itself as the medium. There's a tradition here that goes back to early net art and the demoscene, but what's happening now feels more politically charged.

New York-based artist and educator Everest Pipkin has written and spoken extensively about what they call "software as a site of care"—the idea that how you build a thing reflects your values, and that most commercial software is built with values that prioritize extraction over people. Their generative text works and small-scale tools are built as deliberate counter-examples: slow, weird, open-source, and honest about what they are.

That ethos is showing up in installation work too. Think of pieces that let gallery visitors interact with a simplified version of a content recommendation engine—one where the logic is exposed rather than hidden. You can see exactly why the system is showing you what it's showing you. And once you see it, you can't unsee it.

Hacking the Feed From Inside

Some of the most provocative work operates inside the platforms themselves, using their own mechanics against them. There's a long history of tactical media here—culture jamming updated for the algorithmic age.

Activist collectives have experimented with coordinated behavior designed to confuse recommendation systems—deliberately engaging with contradictory content to muddy the targeting data platforms hold on participants. It's not a perfect solution to anything, but as a gesture it makes a point: these systems are gameable, which means they're not the objective arbiters they're presented as.

Other projects flood data systems with noise. Artist and researcher Joanna Moll has done extraordinary work quantifying the carbon footprint of internet infrastructure—turning abstract server statistics into something visceral, like the number of trees you'd need to offset a single day of Google searches. That reframing is itself a kind of intervention. It makes the infrastructure legible in terms we already understand.

Why This Matters Right Now

There's a reason this work is proliferating in 2024 and not 2004. The systems have gotten more powerful, more pervasive, and more consequential. AI-generated content is flooding the information environment. Automated hiring systems are making decisions about people's economic futures. Social platforms are being explicitly weaponized to shape political outcomes.

And the regulatory frameworks meant to govern all of this are laughably behind. Legislation moves slowly. Journalism can document the problem but rarely changes the underlying architecture. What art can do—what it's always done—is shift perception. It can make something feel real that previously felt abstract.

The artists working in this space aren't under any illusions that a browser extension or a gallery installation is going to dismantle surveillance capitalism. But they're doing something important: they're building a visual and experiential vocabulary for systems that have deliberately resisted one. They're making the ghost code visible.

And once you've seen it, you start asking different questions about the spaces you move through every day—even the ones you can't see.