I agree that the founder effect is a significant factor, but I also have another (more crackpot-ish) theory. I think some disciplines, like algebraic geometry, are harder to pick up in a traditional classroom setting, or out of a book. In practice, they are more often learned like a language, through repeated exposure and watching other people do it. Of course, to some extent this is true for most disciplines, but I think its more true for algebraic geometry than for analogously hard disciplines. If you grant my point, then there would be an over-representation of these disciplines on the internet, because they are looking for explainations and intuition unavailable in technical contexts. I think a similar statement is true for category theory, for instance.

I agree that the founder effect is a significant factor, but I also have another (more crackpot-ish) theory. I think some disciplines, like algebraic geometry, are harder to pick up in a traditional classroom setting, or out of a book. In practice, they are more often learned like a language, through repeated exposure and watching other people do it. Of course, to some extent this is true for most disciplines, but I think its more true for algebraic geometry than for analogously hard disciplines. If you grant my point, then there would be an over-representation of these disciplines on the internet, because they are looking for explanations and intuition unavailable in technical contexts. I think a similar statement is true for category theory, for instance.