I'm pretty suspicious of claims that some technology or social trend makes general epistemics a lot better but elite epistemics a lot worse, or vice versa.

First of all, null effects are probably pretty common.

Secondly, for things that do have a significant effect, I think the following are far more likely:

- some technologies/trends uplift the general public a bunch (eg public education, Wikipedia) and elites don't "need it" (but still benefit from it) and are snobbish.

- some technologies/trends downlift the general public a lot (eg social media, shortform video) and also harms elite epistemics, but less than it harms public epistemics.

As opposed to a model where there's a real tradeoff between significant elite epistemics uplift and general epistemics downlift, or vice versa. Just too easy for elites to delude themselves into thinking good things are secretly bad, or bad things are good/not a problem.

The main exception I can think of is siloing/assortative matching. There the story for differential and reversed impact is relatively clear.