The echo chamber is a feature, not a bug

We talk about echo chambers as if they were weather — something that happens to us, an unfortunate side effect of scale. That framing is comforting and wrong. The loop is not an accident. It is the optimization target, dressed up as a convenience.

Every feed is tuned for one number: time spent. The fastest way to grow that number is to show people more of what already moved them. Agreement is sticky. Mild outrage is stickier. A view that genuinely challenges you costs attention to process, and attention that gets spent thinking is attention that didn't get spent scrolling.

So the machine learns to round you off. It finds the version of you that engages most and serves that version back, slightly exaggerated, every day. Over months the exaggeration compounds. You did not walk into the chamber; it was built around you, one tap at a time, from materials you supplied.

The fix isn't a better algorithm — a better algorithm optimized for the same number lands in the same place. The fix is choosing a different number, or refusing to let a single number decide what a conversation is for. That's a product decision, and an editorial one, long before it's a technical one.

Which is the uncomfortable part. The echo chamber is working exactly as designed. The question was never whether we could close the loop. It's whether anyone is paid to leave it open.