Aiaiai art

Since 2024, our social media feeds have been bombarded with more and more AI generated nonsense. Shrimp Jesus, muscular oranges (U Din Din Din Dun), or monkeys inside bananas (Chimpanzini Bananini) have been parading across our screens. This AI slop is generated content with, arguably, no other purpose or meaning then grabbing our attention. Jason Koebler, journalist and cofounder of 404 media, reported that companies like Meta are paying content reators in the Global South[1] to create AI slop to make users scroll as long as possible. He warned that this AI slop can lead to a ‘zombie internet’ where “a mix of bots, humans and accounts that were once humans but aren’t any more mix together to form a disastrous website where there is little social connection at all.”[2]

But while this ‘zombie internet’ looms on the horizon, it’s not necessarily overtaking all corners of the internet. The monkey inside bananas and muscular orange mentioned above are characters in the Italian Brainrotuniverse, an AI phenomena that became popular in early 2025, showing generated hybrid creatures coupled with Italian sounding phrases and nonsensical narratives. While AI slop images and videos often stand alone, brainrot is a universe of strange and surreal characters. Users themselves are also constantly adding newcharacters, pointing to a growing and collaborative gesture. According to writer Al Hassan Elwan, this signals an important distinction: where AI slop alienates us from each other, brainrot brings us together.[3] Some even argue that Italian brainrot is its own artistic movement[4] or an internet native folklore. In her article ‘Why brainrot gives me hope for humanity’ writer Rina Nicolae writes “What’s happening here is the creation of an internet native folklore, collaborative myth making playing out over weeks and months, rather than generations.”[5] On the surface, Italian brainrot characters may seem absurd, but these hopeful perspectives open up possibilities to how AI tools can be used. Even in online environments that are increasingly shaped by machines that alienate us from one another, we still urge for connection.

The brainrot universe is the entry point into our online generative AI landscape, but there is so much under the surface. Whether you’re creating images or videos at the click of a button, cloning your voice, or generating an essay within a matter of seconds, the ways in which generative AI has infiltrated our daily lives is often met with extreme reactions. To some, it’s like a magic box that’s spitting out polished illustrations or eloquent text without the need for the traditional skills, time, and handwork it takes to develop them. To others, it’s more likea black box. One fuelled by massive data extraction, heavy energy use, and the threat of automating not just human labour, but creativity itself.

This tension plays out clearly in education. While students are already using AI tools to draft essays and brainstorm ideas, many educators have responded by banning them completely. They are understandably worried about plagiarism, a loss of critical thinking skills, or are simply unsure about how to integrate these tools into their teaching practices and curriculums. But is exclusion the most constructive response?

Whether we like it or not, AI is already deeply embedded in our digital lives. From recommendation systems to facial recognition technologies, customer service bots, and algorithmic feeds, we’re constantly interacting with machine intelligence, often without even realising it. Billions are being invested by Big Tech giants and governments around the world in accelerating these systems. Generative AI is just the most visible tip of the iceberg. With all these accelerated developments, and a growing user base, ignoring AI just isn’t realistic. But that doesn’t mean we have to accept it as it is. The question isn’t whether or not to use it, but how. How do we shape our relationships with these tools on our own terms and find ways to use them to create frictions and new possibilities? This is the central question that Ai Ai Ai: A Hands-On Guide to Playing, Failing & Tinkering with Machines tries to answer. Taking shape as a physical and digital DIY guide with accessible and fun exercises, Ai Ai Ai teaches (future) creatives how to tinker and experiment with AI technology using a collection of recipes prepared, or inspired, by the work of artists working with AI.

It’s interesting to compare this current stage of AI development to the invention of photography. In the beginning, the camera was simply a tool. It was an efficient way to capture people. Few could have imagined that one day, people would be paid to be photographed. It was only when people began to see the camera as a medium for human craftsmanship—something through which they could express themselves and develop a personal visual language—photography emerged as an art form.

As machines create based on existing knowledge, letting the machine do too much of the talking often leads to an overload of sameness and mediocrity. Due to its training on massive amounts of data sets, large language models are ideal for generating content within known formats—iterating on existing ideas. While two years ago we were surprised by glitched results—which you could see as the machine’s own interpretations—the outcomes are now becoming increasingly smooth and predictable. If we want something new to emerge from the machine—something unexpected, something strange, something that lives on the fringes—we need to resist settling for the first result and find new and collaborative ways to use these tools and technologies. Real creativity with AI requires experimentation, friction, and play. AI practitioners have much to learn from AI artists, who have often been trained in critically engaging with the tools they use. Asking questions such as “How does the tool work?”, “What kinds of value systems are embedded in it?”, or “How might we misuse or subvert it?” opens the door to more meaningful and creative uses of AI. 

So how are artists working with, misusing, and generating new work with AI tools and technologies? This is what we share in Ai Ai Ai.  

  • Eryk Salvaggio, an artist working with AI and one of the contributors to Ai Ai Ai, does not see his relationship with AI as a collaboration, saying “I have no choice. My data is analysed, my picture is taken at the traffic light, the tools are integrated in the university or in the workplace.” Instead of collaboration, he tries to  antagonise—and analyse—back.[6]
  • In their Ai Ai Ai assignment, K Allado-Mcdowell, a pioneer in AI literature, highlights how the design of AI tools shapes the writing process. Early AI tools offered suggestions during writing, allowing for a more collaborative, back-and-forth experience. In contrast, newer chat-based tools are designed more like conversations, often leaving more of the creative work to the AI. 
  • Artist Salim Bayri, in his assignment, asks us to study how tech companies want us to use their tools and then find ways to use them badly, in ways that they’re not supposed to be used, and do it over and over and over again. It’s in these glitches, collaborations, and endless misunderstandings that new work emerges. 

We are delighted that the artists were willing to share an exercise in this guide, giving us a sneak peek into their creative process. In an age of AI, where outcomes can be generated instantly, it’s important to come together, sharing knowledge and value experimentation over the final result.

Now it’s your turn to open this guide, explore the assignments created by artists and inspired by the work of artists using AI tools in their practices. Get your hands a little dirty and get started.