It’s more about recognizing what kind of biases the user will likely be approaching a given topic with. .ml users obviously tend to skew towards viewing world events through a very specific lens, which then goes on to color how they interact with the rest of the fediverse.

The funny part is that when you’re a .ml user on .ml, it looks like smooth sailing and calm waters. Discussion is civil and friendly, and dissent all lands within a very specific Overton window. But once you start digging deeper, you realize it’s because the entire instance is extremely heavily moderated. It looks like everyone agrees because dissent outside of that tiny bandwidth quickly gets silenced. So as long as .ml users stay on their instance, things tend to be pretty chill.

This ironically leads to friction when those .ml users end up on external instances, because the .ml admins can’t moderate those as heavily. Plus those external users have likely seen and/or experienced the heavy-handed moderation firsthand. So they immediately assume .ml users are complicit and intentionally staying in their bubble, instead of simply being unaware.