I understand the distinction. We disagree on what counts as "good content."

You're assuming that because something came from Reddit and was submitted automatically, it's inherently lower value. I'm not.

And again, this isn't a copy-paste script blindly mirroring Reddit. There are multiple layers of curation: score thresholds, freshness limits, duplicate detection, community-specific rules, content filtering, and AI evaluation. Relevance to the specific community is literally one of the things being evaluated.

Could that filtering make mistakes? Absolutely. So can a human moderator or someone manually choosing posts. If particular posts are bad, downvote them, remove them, or tell me the filtering needs adjustment. That's useful feedback.

But there's also a chicken-and-egg problem with "I prefer conversations that are already happening." Someone has to start those conversations. If most users prefer commenting rather than submitting, where are all those conversations supposed to come from?

That's the 90-9-1 problem I've been talking about. Most people consume, fewer comment, and very few actually submit content.

And the results matter. Lemmy has hundreds of thousands of registered accounts, yet only tens of thousands are monthly active, and many communities remain nearly empty. We can't simultaneously say "people prefer joining existing conversations" and then treat supplying things for them to discuss as inherently harmful.

I'm completely open to arguing about quality, volume, thresholds, or relevance. Those can all be improved.

What I reject is the idea that automated = spam and manually submitted = good. Quality should be judged by the content, not by how many mouse clicks it took to submit it.