But my B.S. detector goes off when I see precise statistics, for a bunch of reasons. How well are these self-appointed watchdog organizations monitoring all of Twitter? Users send, on average, 500 million tweets a day. Users sending 3,900 racist tweets in a day is a bad thing, but that would still represent 0.0000078 percent of all tweets. No wonder the average user isn’t noticing it. For perspective, the U.S. has 24,000 homicides per year out of a population about 330 million people, meaning 0.000072 percent of all Americans are murdered in a given year. In other words, racist tweets make up a smaller share of all tweets than murder victims make up a share of all Americans. …

And how much does it matter if someone with only a few followers is vomiting his rage into the void? How much of this is worth worrying about on a platform with block and mute features? The reason you see so many horrible comments on social media is because they offer people a combination that was previously rare in human experience: anonymity and the ability to reach a wide audience. This allows people to express all of the taboo, controversial, or antisocial thoughts that would usually bring negative consequences if expressed explicitly in offline life. It’s a formula to bring out people’s inner jerk, and it’s more or less baked into the cake of the user experience. The only factor that might mitigate it would be eliminating anonymous users, but then a lot of people wouldn’t want to use Twitter.

But all of that complicates the preferred narrative: “Elon Musk bought Twitter, and the hate-mongers took over.” Because Twitter was such an online earthly paradise before then.