VJayBombs has 300,000 followers across Instagram and TikTok for its guerrilla protest art across Los Angeles. The collective makes work that is explicitly anti-algorithm in content: it targets Trump, ICE, and the apparatus of state power whose funding Congress just approved to the tune of $70 billion. But the work distributes itself through exactly the platforms it aesthetically critiques. This week Phoebe Bridgers did the inverse: she ditched the internet to hype her new music, going offline to generate online attention. Both moves confirm the same thing: the algorithm has become the air. You cannot breathe outside it, only position yourself strategically within it.
Street Art's Distribution Problem Is Now the Content
The historical logic of guerrilla street art, from the Situationists through to Banksy, was that the street bypassed institutional gatekeepers. The wall was the distribution. VJayBombs still uses the wall, but the wall is now a content input for the feed. The protest is documented, uploaded, and optimized for engagement before the paint is dry. This is not a critique of VJayBombs specifically. It is a description of a structural condition that the Atlantic's analysis of white identitarian politics actually illuminates from the other side: when political movements of all ideological stripes route themselves through the same engagement-maximizing infrastructure, the infrastructure shapes the politics more than the politics shape the infrastructure. Kyle Chayka's work on algorithmic homogenization is exactly relevant here. When protest art and its targets distribute through identical platform logic, the platforms win.
The Virality of Anti-Virality
Bridgers' move is the most clarifying data point. Refusing the internet, done correctly, is itself internet content. Her presale generated enormous search interest precisely because of the refusal. This is the loop: resistance aesthetics go viral, virality funds the resistance, the resistance performs its anti-virality, and the anti-virality goes viral. A 2026 arXiv paper on generative models eroding human temporal learning through market selection offers the most useful framing: what gets selected is not the most authentic or most oppositional content, but the content most legible to the selection mechanism. Street art and folk singer anti-internet stunts are both readable. The unreadable stuff is what actually escapes. TurboFund's spotlight on flipping screen time from passive to active is trying to build against this same dynamic, betting that intentional engagement is a product category.