The idea that 99.9% of content will be AI-made is very reasonable. When you consider that a lot of AI writing and images is already indistinguishable from human-made content (see scientific studies in the article), we are faced with some scary ideas, such as:

1. AI may be even BETTER than humans at generating interesting and engaging and thought providing and entertaining and addictive art, movies, writing, etc.
2. The AIs will make virtually unlimited amounts of content instantly (for virtually free).
3. The AI's output in terms of ideas, societal values, etc., that they promote will be controlled by perhaps governments, dictators, corporations, etc., that are in control of the AI.
4. The article asks, can we find a solution? Some way to identify content as Human-made?
5. This solution would allow people to search for human-made content only. It would allow them to be exposed to more ideas that fall outside of the officially sanctioned version of the truth. It would also promote human beings to keep being creative and keep discussing ideas and progressing and growing intellectually and creatively as a species.
6. One perspective is that human content will be rarer and it might be perceived as "REAL" and therefore more valuable like a real diamond is more valuable than a synthetic one even when the synthetic one actually looks more perfect and beautiful. Since there will be a perceived greater value for human content, perhaps this will inspire solutions that allow people to search for human content and trust that it is not written by a bot.