In Why do I suck, Scott Alexander is analyzing why his earlier articles (from 2013 - 2016) were much more popular than his recent ones.
I would like to highlight his insight into the perception difference between a small blog and a big one:
If you have a small blog, and you have a cool thought or insight, you can post your cool thought or insight. People will say “interesting, I never thought of that before” and have vaguely positive feelings about you. If you have a big blog, people will get angry. They’ll feel it’s insulting for you to have opinions about a field when there are hundreds of experts who have written thousands of books about the field which you haven’t read. Unless you cite a dozen sources, it will be “armchair speculation” and you’ll be “speaking over real academics”. If anyone has ever had the same thought before, you’re plagiarizing them, or “reinventing the wheel”, or acting like a “guru”, or claiming that all knowledge springs Athena-like from your head with no prior influences.
I try really hard to block or ignore these people when I spot them, but they do a little bit of psychic damage each time.
I wonder how much of that is simply bigger audience = worse audience, given regression to the mean. Basically: when your blog has 100 readers, it is not difficult for this 100 to be “reasonable people” (from you, the blog author’s, perspective), but if you have 10 000, then it is much harder: as the size of a group grows, the more it tends to resemble the population at large. The population as a whole contains a whole lot of people who suck! (This seems universally true, regardless of what group you identify with.)
Another perspective on the same idea: out of 10 000 people, you’re much more likely to annoy / step on the toes of somebody by something you write than out of 100 people. 10 000 people will also likely contain some trolls and griefers, who just tend to make things worse.
The other possible explanation is authority. A blog with few readers has little to no authority on any subject, and therefore is largely safe from the kind of scrutiny that having authority (even informal!) invites. So you can get away just expressing your opinions in a way that makes sense to you. Doing so when you have a large audience is risky, expectations are higher, and the whole thing becomes a lot more high-stakes.
Taking this further, what does this mean for official authorities like the CDC or the FED? Obviously that’s kind of as high-stakes as it gets, but beyond that, they are very constrained in what they can actually say if they want people to actually take them seriously. They can’t just say things that make sense to them! It needs to be as unambiguous as possible, supported by research and data.
As an aside, I find it interesting that one of his plausible explanations is just that media overall is less bad today than it used to be. Not often do you see this proposed!
Nowadays I think there are many good science bloggers, and the media has gotten embarrassed enough times that it will sometimes run a take by someone who knows what they’re talking about before publishing it.
In the same way, I see fewer people outright denying the existence of genetics, totally failing to understand AI risk, or utterly bungling basic concepts in risk and probability.