so the UK influx brought in some big-name TERFs like https://bsky.app/profile/naomicunningham.bsky.social

this being Bluesky, they stuck an "⚠️ Intolerance" label on the account… and that's it. that's the Bsky promise: they basically don't moderate. stick to politely threatening people's existences within the ToS and you're good.

#Bluesky

pretty much running like i predicted https://princess.industries/@vyr/statuses/01HP2T5XWARBB7DQHN81Q8H4A9

there's also been an anti-trans troll campaign going for the past few days, but i think that one is a few idiots using manually registered accounts, and saying shit that definitely violates the ToS. still, Bsky doing a remarkably bad job of catching them.

#Bluesky

and here's the head of Bluesky T&S making a long speech justifying doing fucking nothing, which he was hired to do, while very carefully avoiding naming anyone who might be harmed by this approach: https://bsky.app/profile/aaron.bsky.team/post/3l2owilimci2i

The inversion principle is a powerful tool I use to shift my perspective and uncover blind spots. Consider this example: "I was repeatedly rude to this person because they are bad." Here, user A labels themselves as good, justifying rudeness to user B, whom they've labeled as bad.

People who agree with user A might think it's acceptable to be rude to user B due to their "bad" views. To check for bias, flip the model by inverting the users. Is it acceptable if user B does the same to user A? If not, there's a bias favoring one group over the other.

T&S policies should focus on behaviors (e.g., treating others with respect), regardless of the direction of rudeness. Otherwise, policies could be weaponized to favor one group over another repeatedly.

While bad faith actors exist and context matters, most cases involve two people trying to spend time online. Often, one person believes their abuse is justified based on the other's ideas or identity.

I've kept labels neutral to avoid polarizing the discussion with conflict, identity, or political associations. The goal of this self-questioning is to ensure impartiality when consulted on policy violations by moderators.

This principle helps me maintain neutrality in decision-making. By inverting perspectives, I can better understand and address potential biases in our approach to content moderation and user interactions.

I wish I could find the Twitter thread where I first saw it for credit, but I could not find the post from several years ago. I look forward nonetheless in learning and applying new ideas from Bluesky threads in the future!

#Bluesky

@vyr incidentally, this type thing is what I hear every single time someone tells me they're definitely qualified to talk about moderation because they worked in "Trust and Safety" at some kind of social media company