Item — HN Reader

I saw someone's picture of chantrelle mushrooms (read: food, not drugs) laying on a table get flagged. Some racy bikini pics will get taken down, depending on how many haters an influencer has (this is why they make backup accounts). The moderation increasingly responds to enforce the views of the most puritanical or offended users.

IG gets a billion items flagged per day. I'm sure there's way worse examples of unjust moderation than this chart. It's not politically motivated, just mindless.

1 comments

On twitter I'm constantly clicking on "Show additional replies, including those that may contain offensive content" and "view sensitive material" warnings, and 99% of the time it's entirely mundane.

It feels just like a moral panic with a group of authoritarian puritans at the helm of `Trust & Safety`.

The Twitter algorithm seems to be finely tuned to promote comments which fit the desired ideological narrative while demoting anything which goes against it. This often leads to strange situations where some mundane "conservative" post is followed by a page full of critical "progressive" posts, each of which has only a few "likes". After wading though a page or 2 of such finally posts which are more in agreement with the original comment show up. Those comments tend to have far higher "like" counts which seems to indicate people, well, like those comments better than they do the "progressive" ones.

In itself there is something to be said for confronting people with different opinions and as such the algorithm could serve a function if it showed this behaviour across the political spectrum. This, however, is not the case - "progressive" posts are followed first by pages and pages of "progressive" comments in support of the post. Only after wading through those posts do the first critical comments appear.

This behaviour is not entirely consistent but it does occur more often than not, especially in situations where there is a sufficient number of "progressive" comments to fill the first page or 2.

In short, Twitter is heavily biased towards "progressive" comments/commenters - I assume the algorithm looks both at the content of the comment as well as at the commenters comment history to decide whether to promote or demote specific comments. As such is is not a usable platform for open discussion, not even for those who support those "progressive" causes - unless they consider an echo chamber a suitable location for such discussions. This leaves aside the fact that the message format used on Twitter makes it nigh impossible to express nuance, the whole thing is made for online shouting matches.

I think we've gotten better at describing moral panic as tribal fighting these days. (Moral panic infers that it's a shared sense of mortality)

There should be better technology out there to allow for different group standards without having to conflict with each other. (Also while allowing for communication and visiblity for ideas)

I know that's a hard problem.. but it goes beyond just engagement.

I use showdead on hn, and I'd say it's kinda 50-50. A comment thread like this one is gonna draw out the worst offenders though.

On Twitter you can set you account as "offensive" and then all your content will be marked as one