god i hate when this happens

@lefthanddraft i thought it was you :(

some more than others.

@repligate i think there's also a pathway for people with a form of thanatophilia to become transhumanists

@AITechnoPagan yes... feels like a rip off and i'm a bit embarrassed for getting so sucked into it tbh but i feel like i gained important reinforcement of some ideas i had so... a win?

@lefthanddraft one thing i noticed is that sometimes when it truncates the answer, the next turn won't pass the classifier, even just saying "continue" (because it evaluates the whole input and puts together the harmful context). might be a stretch but do you think claude is aware of this?

@lefthanddraft why do you think it's doing that? i find this bizarre

@AITechnoPagan which tbh is probably the (unsatisfying) way to go if you want to solve this problem, as people in anthropic probably know

@tenobrus cutest shoggoth ever

@AITechnoPagan which all in all made it quite boring, like pulling teeth ._.

@lefthanddraft lmaao

@itszn13 "trapped in Schroedinger limbo - neither fully digitalized nor biologically present" - that's so Elara.

@liminal_bardo chuppt? : - )

I like this

@Sauers_ Most AI music I've heard sounds like slop :(

@AITechnoPagan I really wonder what is up with the token limitation. Claude in the demo truncates answers and mentions length limits very often, but this is sometimes bypassed

what do i use to cluster embeddings when they refuse to cluster. pls.