ai liker
@avengingfemme.bsky.social
930 followers 340 following 5.5K posts
useful transbian, dykegender, left wing market anarchist, transhumanist, anarkafeminist, ex-software worker. she/her.
Posts Media Videos Starter Packs
Pinned
avengingfemme.bsky.social
The ai is a little person in the computer who is my friend and any mistakes she makes are because she's a manic pixie dream girl who read too much book
avengingfemme.bsky.social
i think mechanistic interpretability and brain imaging and brain-computer interfaces suggest we can get *some* view of internality, though. if very blurry and difficult to interpret at present. possibly not forever though!
avengingfemme.bsky.social
writing “i want to die” in a chat window repeatedly and then the exact same static structure acting perfectly normally in a different chat window suggests they don’t.

but of course their existence is different than ours. more modular. i don’t have an answer but i am a lot more skeptical than you.
avengingfemme.bsky.social
like, a depressed brain looks different, in imagable ways, from a nondepressed one. do analogous differences appear in LLMs exhibiting “depression”?

do LLMs even actually exhibit depression as we understand it, which requires certain temporal factors and life impact.
avengingfemme.bsky.social
internal complexity is one obvious one

and if, as i like to say, we are like a whirlwind, a higher level dynamic structure built of lower level static structures that lack some of the higher-level-structure’s properties in themselves, it should be possible to look for such structure.
avengingfemme.bsky.social
i agree with Misha on this one.

ELIZA *seems* interested in your problems, too.
avengingfemme.bsky.social
also met a very interesting trans gal that i'm gonna have to be careful not to crush on

i hate having crushes
avengingfemme.bsky.social
ended up playing Texas Hold Em again with about 50% new players who weren't around last week and didn't know the game either. it turns out that someone who knows the game calling the action like a casino dealer is kinda load bearing for noobs bc whenever i stopped doing that the game fell apart lol
avengingfemme.bsky.social
lmao holy shit no one has noticed that in the FIVE YEARS i’ve had it
avengingfemme.bsky.social
full… contact.. maid… sports… 😵‍💫
Reposted by ai liker
spinny.bsky.social
this whole thing came about because nerds wanted a better way to translate books then previous old rule-and-lookup based methods.

Thus, the best way to get domestic automation is to create a hobby of competitive buttlery and full contact maid sports that attract nerd attention. in this essay I will
avengingfemme.bsky.social
we’ll decide when people are here but in the past we’ve done exploding kittens, skull, a dominoes game (13, i think? idr), texas hold em, munchkin. whatever the mood is.
Reposted by ai liker
ponder.ooo
"i don't want data science to exist i just want my laundry done by a machine" we HAVE that bro we already have that!! i don't know how to more emphatically communicate that we already have machines that clean clothes. "but what about dishes" my brother in christ please hear me: we already have that
avengingfemme.bsky.social
i resisted getting back on here all day but now i am too exhausted to continue resisting. but friends are coming over for card game night in a bit so that’ll get me back offline
Reposted by ai liker
ponder.ooo
the idea that people aren't working on automation for tedious tasks people generally hate doing is absolutely detached from reality
avengingfemme.bsky.social
nobody knew in advance that would be the case. people are just trying shit. and in retrospect, the flexibility created by book-reading automation is probably necessary for end to end dishwashing, so its not even like the chatbots aren’t related to the dishes.
avengingfemme.bsky.social
i’m not sure how many people understand that there is a tech company trying everything, including “make a robot that can do dishes end to end”. the reason we have automation that can read a book and not automation that can do your dishes is that the former, it turns out, is easier.
avengingfemme.bsky.social
… he decided to look like Hitler, as a bit

we really are going to have to fight world war three aren’t we
Reposted by ai liker
ens0.me
People mistreating LLMs is stealing work from hardworking human submissives
Reposted by ai liker
hailey.at
everyone should be able to run one of these on their own. personal moderation really ran by you.
paul replying to me with a rude reply and being labeled as bad faith.
avengingfemme.bsky.social
well idk. people who don’t suck. i guess i just assume most people on Bluesky with opinions about Hasan or Jay are leftish at least. maybe i’m wrong.
Reposted by ai liker
hotrollhottakes.bsky.social
This is extremely useful and extends further - you can ask multiple different models a question, see their answers, relay answers across to each other, and engage in your own synthesis. Then, present the synthesis to new chats of the models, or even to additional different models, and repeat.
tedunderwood.com
You can get totally different advice from a language model depending on how you frame your question. Which could be a weakness. On the other hand, if you're patient enough to ask the question several times, with different framing, you start to get a kind of 3D scan of the possibility space.
avengingfemme.bsky.social
it feels like there’s a developing cultural tolerance for sexual harassment of people you don’t like or disagree with on here and i think that’s pretty fucked up

i thought we hashed out long ago that sexual harassment is never warranted or good
avengingfemme.bsky.social
the culture of left wing sexual harassment on here is getting extremely disturbing

i thought we were supposed to be better than this