✨🧋Tea🧋✨ (Comms Open) (@tea-dragonz) — Bluesky Profile

4 months ago

reminder to click here: @moderation.bsky.app
and set these 2 settings to off to avoid posts like this being hidden from you

155 25 6 2

4 months ago

Stream redeemed catpic.

Bear was giving me horrid side eye.

116 13 1 0

4 months ago

Soso and fam are saving up to start a new life in a safe place where her little sons can grow up in peace.

Please continue to support them, they need all the help they can get.

@mommunism.bsky.social 🩷🩷🩷

77 91 3 14

4 months ago

#spamtennatober
#spamtenna
Did I make this as a continuation of the last one? YEP

Day 7: Prophecy
Day 8: Smoke Break
Day 9: Maintenance

54 8 0 0

4 months ago

Lots of families are seeing their donations go down since the ceasefire which is understandable but also they’ve lost everything and are hoping to rebuild or at least give their kids a stable home through the winter- if you can help ❤️

62 89 2 0

5 months ago

Woke Bear up from a nap to grab a pic. He wasn't happy

Another stream redeemed cat pic.

117 15 1 0

5 months ago

Mizu is a glorious small kitty.

Stream redeemed another cat pic.

122 16 3 0

5 months ago

THIS TOOK SO LONG IM SORRY-

here's my take on #Spamtenna UT au and I love them so much guys-

1/5

74 16 2 0

6 months ago

Mizu wanted to be on my pillow.

Stream redeemed another cat pic.

130 11 1 0

6 months ago

#Art #Abstract #humanemotions

552 51 9 1

6 months ago

Screenshot of tumblr text post by fourohfourlifenotfound August 24 part 1/3:

"If we are ever going to stop this, then you guys need to start learning why this happens.
And it's not always because some executive made a decision to favor a TERF. It's because the technology is broken in a way TERFs have learned to use to their advantage.
So content moderation is a HUGE task. It is incredibly inefficient to manually monitor all user reports-- when you "flag" something as spam or inappropriate or 'inciting harm' (what Blueskye is saying the banned users had in their posts). So what actually happens is that the bulk of it is reviewed by machine. The precise "how" is a bit of a black box (if you say exactly, it makes it more vulnerable to being bypasses), but we generally know it's a bit of machine learning ("ai", but not he chatgpt kind) and a bit of automation. Generically this gets referred to as an "algorithm" (which is a meaningless word I don't have time to unpack here).
The TERFs have a formula. A battle plan, if you will. I've been watching them use it for almost a decade. And they play social media sites like a fiddle every single time."

Text post part 2/3:

"The way it works is that they coordinate attacks on targets they dislike. (Usually trans people, but ace discourse is a notable example of targets not necessarily fitting that demographic.) They send links to each other of posts to report. They'll flag it for whatever the nearest thing they can is. The blueskye users were wishing ill to JK Rowling, so "inciting harm" was the excuse selected. "Inappropriate or sexually explicit content" is a common selection for other trans content.
Then what happens is the content moderation "algorithm" sees a huge influx of reports for a certain post or user. And instead of making a human manually review all of that, it automatically initiates a ban or takedown. And if the "algorithm" has word censorship built in, it will begin to auto-flag for things containing words in posts frequently flagged, so it begins to auto-flag the word trans itself.
This has been happening for almost a decade. There was a trans youtuber who proved TERF hater group chats flagged his videos so much that it started auto-demonitizing identical videos with the word trans in the title, but leaving ones without the word alone. I strongly suspect this is why we see huge amounts of trans women banned for no reason on tumblr. And this is clearly what is happening now."

Text post part 3/3:

"How do we fix it? There's no easy solution. All the things about the "algorithm" that make moderation easier for the social media platforms are the exact reasons it's so difficult to defeat. It's hard to teach a machine what's a genuine flag and a malicious one. Again, it's been almost a decade since the issue was raised with youtube, and there has yet to be a solution for rooting out malicious coordinated flagging.
But here's one thing that can help: get TERFs off of social media. Stop letting them bait you into petty fights. Learn how to identify them. Block and report them. "

Response by drtanner:

"Blocking TERFs and other assorted bigots is especially effective on bsky, because the block function there is nuclear. In short, if you block someone, not only can they not view your account or any of your posts, but any replies they made on your posts will no longer be viewable by others, even if they haven't personally blocked that user themselves. If they quote you, you can even decouple your post from their quote. In short, blocking someone on bsky produces mutual hiding of content, and it's extremely effective at starving these cunts of oxygen. Bigots who can't convince people to fight with them and whose posts don't gain any traction because they're blocked by everyone and their dog leave the platform, i.e. Jesse Singal.
It works, people just need to remember to do it more. Preemptive blocking of dickheads before they start shit with you is the best way to prevent things like this from happening. "

Very interesting algorithm analysis, which explains the repeated problems with content moderation getting in the way of social justice

tldr: it's always good to block a TERF or a white supremacist or any other type of fascist

original post: www.tumblr.com/chiptrillino...

1,460 856 7 48

6 months ago

I’m starting commissions!!

Here’s some examples of my work!!

Here’s the Carrd where you can find all the info!

74vinium.carrd.co

1 1 1 0

6 months ago

I woke Bear up for a quick pick he was not happy about it.

Stream redeemed cat pic.

128 10 6 0

6 months ago

My family is suffering after losing our home and livelihood because of the war. We are left with no shelter and no food. I humbly ask for any urgent support to help us survive. 💔🙏

chuffed.org/project/1251...

25 27 3 6

6 months ago

2/2

18 2 0 0

6 months ago

Made this forever ago trying a new comic style concept but kept pushing it back to post post but you take child support AU

1/2

#spamtenna

41 9 2 0