cc @internet.bsky.social

no? when i like a post i am not making a decision that it is good for everyone. i am expressing that i personally like it. other people can interpret that however they want and their interpretation is their responsibility

yeah exactly, i don’t want to be held hostage by those people, i want to express my preferences (that their content is bad)

yeah “people might get mad about other people’s preference signals” is not a compelling argument to me because the people in that group already do that and will do that regardless and i don’t want to be held hostage by them

just make them public records should be the default, this is supposed to be an open and interoperable platform

what are the problems with making negative preference signals public that don’t apply to positive preference signals and why do they outweigh the loss from not being able to express them?

i just don’t think there’s a way to square the circle of “have people make mod decisions on my behalf” and “those people are not trusted to make mod decisions on my behalf”

downvotes would be a great step forward here - if they were part of the protocol - but since they aren’t i don’t think they will work well, making them private cuts the eco off from doing product/affordance discovery which is bad bc that seems like a current capacity constraint of the org

if content moderation shaped affordances continue to be the only user accessible affordances for downregulating content the spiral of legalistic policing will continue. bluesky does not need better community moderation tools, it needs tools for expressing preferences

this is a huge mistake that will lead to further misery. the problem is not going to be solved by more affordances for a legalistic and justiciable frame of content mod, because the core problem is that most bad content is simply bad and i don’t want to see it — but there are no affordances for this

so don’t outsource your decisions on who to follow to random people? just don’t subscribe to giant block lists

claude is good at cryptography

:cheshire_cat_smile:

happy halloween

ayn rand moment tbh

banger

that’s true! hence why it’s important to explicitly conceptualize as part of the threat model

Every time I see physical attacks on TEEs, I wish someone would put out a position paper on @hdevalence.bsky.social 's "guy with a glock" model
x.com/hdevalence/s...

that’s not an excuse not to add alt text