Algorithmic hustle: a critical political economy of AI slop production and consumption in South Korea

In the vast, noisy landscape of the internet, a strange new phenomenon has emerged: videos that look and sound real but contain almost no human thought behind them. These are not the clumsy, funny mistakes of early computer graphics, but highly polished streams of images, voices, and stories generated entirely by artificial intelligence. They are designed not to teach, entertain, or express an idea, but simply to keep your eyes on the screen long enough to generate advertising revenue. Researchers call this "AI slop." It is a term that describes content stripped of meaning, created in massive quantities to game the systems that decide what we see online. To understand why this is happening, we must look at how digital platforms work. These platforms act like giant marketplaces where attention is the currency. They use complex computer programs, or algorithms, to decide which videos to show you. These programs do not care if a video is beautiful or true; they only care if it keeps people watching. When creators realize that the system rewards quantity and speed over quality, they begin to produce content that fits the machine's needs perfectly, even if it has nothing to say.

A team of researchers from Korea University set out to answer a puzzling question: why has South Korea become the global capital of this phenomenon? While the internet is everywhere, South Korea stands out as the place where this type of content is produced and viewed in the highest numbers. The researchers analyzed data from thousands of YouTube channels and found that eleven channels based in South Korea alone had accumulated 8.45 billion views. This is significantly more than the next highest countries, with Pakistan in second place at 5.34 billion views and the United States in third at 3.39 billion. One single channel, called "3 Minutes Wisdom," had gathered over 2 billion views on its own. The researchers did not find that South Korea had better internet or more powerful computers than other nations. Instead, they discovered that the country's unique social and economic history had created a perfect storm where producing meaningless content became the most logical choice for workers trying to survive.

The study introduces a new concept to explain this behavior, which they call "algorithmic hustle." In simple terms, this is a way of working where people optimize their labor to fit the rules of the platform, even if it means giving up creativity and meaning. The researchers argue that this is not because these creators are inefficient or cynical, but because the system rewards them for it. In South Korea, this takes two specific forms. The first is "anonymous hustle," where creators hide their identities completely. Because having a side job can be stigmatized in Korean society, and because the algorithms do not care who made the video, producers strip away their faces, voices, and names. They become invisible machines, producing content that is detached from any real person. The second form is "combative leisure." This describes a situation where the line between rest and work has disappeared. People are so exhausted by the pressure to be constantly productive that they turn their free time into a second job, using AI tools to churn out videos while they should be sleeping or relaxing.

To understand why this happens so intensely in South Korea, the researchers looked at the country's history of "compressed modernity." This is a term describing how South Korea underwent rapid industrialization and social change in a very short time, squeezing decades of development into a few years. This created a culture that values speed and efficiency above all else, often prioritizing growth over safety or deep reflection. The researchers found that this cultural mindset fits perfectly with the way digital platforms operate. Just as the country once raced to build factories and infrastructure quickly, it now races to adopt new technologies like artificial intelligence. However, the rules and ethics needed to manage these technologies have not caught up. This gap means that people learn how to game the system before they learn how to judge the value of what they are creating. The pressure to keep up, combined with a job market where many people must hold multiple part-time jobs to make ends meet, pushes workers toward the path of least resistance: using AI to make endless streams of content that require no human soul to produce.

The researchers examined three main types of this content to see how it works in practice. The first type uses AI to generate images, such as slideshows of animals or historical scenes, paired with robotic voices reading scripts. The second type uses AI voices to read books or give advice, with simple stock footage in the background. The third and most popular type uses advanced video generators to create short clips, often showing animals fighting or interacting in impossible ways. These videos have no story, no educational value, and no connection to reality. They are purely visual stimuli designed to trigger a reaction. The study found that the more meaning is removed from the content, the cheaper it is to make and the more likely it is to be recommended by the platform. For example, the channel "3 Minutes Wisdom" was so successful that it earned an estimated 4.04 million dollars in advertising revenue in a single year. When the platform eventually removed its content for violating rules, the channel did not disappear; it simply changed its name and started posting videos of video games, keeping its millions of subscribers intact. This shows that the system rewards the format, not the creator.

The paper also highlights a contradiction in how the platform, YouTube, handles this issue. The company has publicly condemned this type of content, calling it "AI slop" and promising to remove it. However, the way YouTube pays its creators actually encourages this behavior. The system rewards videos that keep people watching for a long time, regardless of whether the content is good or bad. The researchers found that YouTube removes specific channels but leaves the rules that created the problem unchanged. As soon as one channel is banned, others quickly take its place, using the same methods to fill the gap. The platform acts like a gatekeeper that occasionally closes a door, but the hallway behind it remains open, and the traffic simply finds another way in. The study suggests that as long as the financial incentives remain focused on views and watch time rather than quality or truth, this cycle will continue.

Ultimately, the researchers conclude that AI slop is not a failure of technology or a sign that people have lost their taste for good stories. It is a rational response to a broken system. In South Korea, the combination of a high-pressure work culture, a history of rapid development, and a digital economy that pays for attention rather than meaning has created an environment where producing empty content is the smartest thing a worker can do. The study argues that fixing this problem requires more than just deleting bad videos. It demands a change in the rules of the game itself. If the platforms do not change how they pay creators to value meaning over metrics, and if societies do not address the economic pressures that force people into this kind of work, the flood of meaningless content will only grow. The Korean case serves as a warning: when the incentives of a digital system clash with human needs, the system will win, and the result will be a world full of noise that says nothing at all.