Making money from generative AI video is practically impossible today. The first creators managed to skim the cream, but the market became saturated in just a few months — and the window of opportunity slammed shut as quickly as it opened.

An illustrative case is the story of Kun Chen, a former engineer at Meta, Microsoft, and Atlassian, which is recognized as an extremist organization in Russia. Using the Chinese market as an example, he clearly demonstrated how short AI films turned into a mass phenomenon thanks to the quality of modern video models. Those who were the first to master film and comic generation earned very well: they released many times more content than traditional production allowed.

But the advantage proved short-lived. The earnings disappeared for several reasons at once.

  • All the pioneers' experience fit into a single text file with instructions;
  • within a few months, everyone learned to make films of comparable quality;
  • productivity grew for everyone at once — and no one had an advantage anymore;
  • a race for survival began, in which creators work more than before the advent of AI.

Chen explains this by the lack of protection from competitors. A rare human skill is expensive precisely because such specialists are hard to find. But the skill of working with AI is copied endlessly, and there is no rarity in it. Therefore, a sustainable business built on creating something with the help of AI cannot be established.

Where creators' money goes

The main profit in this market goes to companies like ByteDance. They control viewer demand and distribution channels, and creators have to buy advertising from them to get traffic. The cost structure is telling:

  • 80% of operating costs goes to advertising;
  • less than 5% goes to AI tokens, that is, to paying for the work of neural networks.

For investors, this means that betting on AI video creators and studios has become risky. The content itself costs almost nothing to produce, but it also provides no advantage. Revenue settles with those who own the audience. There is more content that people like, but the platforms have taken all the economic benefit.

Chen quotes a line from a Chinese poem: "those who raise silkworms are not the ones dressed in silk." He warns that without a fairer distribution of wealth, AI will turn billionaires into trillionaires, and ordinary people will be more exhausted than ever.

Cryptalist analysis: The case demonstrates the classic commoditization trap: when a production tool becomes universally available, all the margin migrates to the owners of distribution channels and audiences. For investors, this is a signal to reassess venture bets on "AI studios" and shift focus to platforms that control traffic and data. A sustainable advantage in generative content will be built not around models, but around brands, distribution, and unique intellectual property rights.