AI slop is often described today as a problem that spoils the internet. In reality, however, it is more a symptom than a cause. It is a logical consequence of an environment we have set up ourselves: algorithms optimised for quick reactions, monetisation tied to volume and attention measured in seconds. In such a system, content that is cheap, easy to replicate and extremely adaptable to trends naturally emerges.
According to an analysis by Kapwing, so-called AI slop videos are estimated to generate $117 million a year. Kapwing identified 278 purely “slop” channels on YouTube, which together gathered over 63 billion views and 221 million subscribers, without human creativity in the traditional sense.
I fully realised this mechanism while listening to an episode of the podcast Colin and Samir devoted to predictions for 2026. That is where the Kapwing data came up, in the context of how attention will continue to be diluted and how the volume of low-cost content will grow.
A few days later a public dispute began to unfold in the Beckham family, and within a single morning the feed was flooded with AI interpretations of Victoria Beckham dancing at Brooklyn's wedding. This moment was also widely covered by mainstream media and became the basis for memes, remixes and AI videos spread across platforms.
This type of content is not a problem in itself. If it is used in context and with humour, it can extend a story and earn its place in the wider conversation. That is exactly what creators have done for years with memes, reaction content or comments on current events. The problem begins when this logic turns into mass production of content with no ambition to offer value or build a relationship.
At the same time, in recent months it has been said more and more often that people on platforms such as TikTok feel less connected to content. TikTok itself openly describes in its creator materials that the key moment of a video is the so-called hook, and that it measures the success of content by whether a viewer stays for at least the first 6 seconds. This threshold serves as the basic proxy for capturing attention.
This confirms that attention today does not work linearly. It is fragmented and based on constant re-capturing. The brain is not looking for a smooth experience, but for a hook. And then another hook. And another. AI slop uses this mechanism perfectly, while also speeding it up.
But it is important not to be naive. AI slop will not lead to a mass exodus of audiences from platforms. Data and user behaviour do not confirm that so far. Rather, a different and less visible shift is emerging: people are staying, but more shallowly. They spend less time with any single creator, bond less and remember less. Attention has not disappeared. It has simply become harder and harder to turn into a relationship.
This is exactly where the current narrative about a return to raw content and long-form formats begins to break down. It is not about nostalgia or a moral turn by audiences towards quality. It is a reaction to an unsustainable model of working with attention. More raw content, more long formats, more room for connection instead of perfection. Not because that is prettier, but because it better matches the human need for connection in an environment that is increasingly mechanistic.
The claim that “depth will win” is not enough on its own. Depth without context and without an entry point does not work. Long-form content is demanding in terms of time and mental effort, and audiences choose it selectively. Those who simply slow down will not win. Those who manage to combine two extremes will: a fast, clear entry point that respects the reality of fragmented attention, and a deeper format that gives people a reason to return.
The financial motivation around AI slop will not disappear. It is too efficient, and the technology will keep improving. The critical question is therefore not whether it can be avoided, but where it has its place in the whole ecosystem. Creators who see it as a tool have a chance. Those who build their entire relationship with the audience on it will probably hit its limits sooner than the algorithms do.
This is not a return to authenticity as an aesthetic. It is a return to owning the relationship with the audience. And that is something no AI slop, however large its volume, speed and reach, will replace in the long run.
Comments