How Twitter Solved One of Its Oldest Problems

The algorithm hasn’t fixed harassment, but it did remedy something else

It’s hard to remember now, but there was a time when one of the central dilemmas Twitter users faced was figuring out who to follow — and who not to. Because the service fed you every tweet in reverse chronological order, each person you followed meant a more crowded timeline, making it harder to keep up and increasing the chances you’d miss the tweets you really cared about.

One reason it’s hard to remember is because that problem has been overshadowed by so many other, weightier ones: harassment and hate speech, state-backed propaganda and misinformation, polarization and filter bubbles. Figuring out how to optimize your timeline can feel like a trivial concern compared to simply staying safe and sane while tweeting.

But another reason it’s hard to remember is that Twitter has more or less solved the problem, albeit by accident, and seemingly without anyone noticing. It’s one of the subtler side effects of a controversial algorithm, introduced three years ago, that aims to serve users tweets they’re more likely to engage with, rather than just the most recent ones from everyone they follow. That algorithm has upended the service in ways that remain poorly understood but have helped to…