banner
lilykiwi.xyz
@lilykiwi.xyz
https://lilykiwi.xyz/
the meowerrrrrrr

Sent from my archlinux
🩷 @emmeowbun.bsky.social 🩷
Reposted
November 8, 2025 at 5:57 PM
November 9, 2025 at 5:07 PM
ubuntu
November 9, 2025 at 5:06 PM
November 9, 2025 at 4:54 PM
fucking microsoft
October 29, 2025 at 8:02 PM
why cant i remove this
October 29, 2025 at 8:02 PM
HUH?
October 29, 2025 at 8:02 PM
out of interest which symbol do you idealize in your profile picture
October 19, 2025 at 3:04 PM
they dont wanna comply to the UK's online safety act (which is an incredibly stupid law) so they just cut service to the whole country
October 18, 2025 at 3:33 PM
chat i forgot about sgez
October 14, 2025 at 2:08 PM
o-oh. how was i supposed to know!!!!!!!!!!!
October 13, 2025 at 4:09 PM
duh you gotta get it wet and hot and take your time with it
October 13, 2025 at 4:07 PM
yr talking about whether ai image classifiers for moderation is good or not and it's not as easy as "yes" or "no" lol.

every person i've seen talking about this is saying "there's a bias here and it sucks"
October 7, 2025 at 9:59 PM
i'd say that about facial recognition ML as well. these systems have huge problems with bias. we cant just go "oh they are fine with no issue :)" or "this is totally wholly wrong and evil!" because it's more nuanced than that lmao. i think its a good use of the tech but said issues need addressing.
October 7, 2025 at 9:51 PM
who outright claimed that ai moderation is bad?
October 7, 2025 at 9:46 PM
anyway jay posting on main "decentralisation accelerationist" along with a shoddy AI poster imminently after the waffles incident should tell u a lot about how staff see this and related issues
October 7, 2025 at 9:42 PM
said decisions mostly revolving around their new(ish) ToS, application of said ToS against the people in their replies "harassing human people for fun". is it particularly difficult to establish a bias against trans people from this current moderation? i dont think it is
October 7, 2025 at 9:42 PM