Well and truly fucked. Cheers! Hmm...I post/reblog about whatever catches my attention (multifandom? sure, why not). Ask is open to anyone and everyone, no need to be shy (particularly if you want to talk about music!).
I don't usually do poetry, but contrapuntal poems have made me curious since I first read one (I think it was one of @two-bees-poetry's) and I wanted to try my hand at it. I was inspired by @fourseasonsblossoms's wonderful Nirvana in Fire contrapuntal poem series as well.
(many thanks also to @fromaliminalspace for being enthusiastic enough about it to make me want to post it <3)
If you've never encountered one, it's a 3-in-1 poem for two voices: I like reading the left side first, then the right, then the entire thing.
Text-only version under the cut.
Lin Shu
you lie; in agony on the blood-soaked snow
you had fire in your heart.
you were a soldier once,
do you remember?
you died.
you do not yet understand
how it will change you
be glad,
you will find out
open your eyes.
destruction is all you can see
a battlefield,
bloody lips and broken skin
a corpse, speaking. we know you
we know you will try.
dream.
within the confines of your mind,
avenge with your blade
the ghosts all around you.
close your eyes
this rage beneath your skin
the monster you’ve become
it’s a promise.
all this pain you’ve endured
someday it will end
the spirits of the dead hold you up in their arms
freezing wind blowing on your exposed skin.
for them
you will stay
you will live.
Mei Changsu
you lie; on your back unable to breathe
illness does not suit you
but it is the choice you made.
bones ground to dust, skin flayed apart
you survived.
the price you paid for this reprieve:
your face, your health, your self.
they will tell you, averting their eyes
that it was too high.
rise
from your bed; it has become
a prison.
this body no longer yours
will fail to achieve your goal.
find another way.
plan
the rise and fall of empires
the death of thousands.
the next day, repent
start over.
the poison in your veins
slowly kills you
you will keep it to yourself:
it will be over soon
just as it began.
the embraces of the living are nothing but
the scattered ashes of broken promises.
alone
you will wait
you will die.
Lin Shu / Mei Changsu
you lie; in agony on the blood-soaked snow / you lie; on your back unable to breathe
you had fire in your heart. / illness does not suit you
you were a soldier once, / but it is the choice you made.
do you remember? / bones ground to dust, skin flayed apart
you died. / you survived.
you do not yet understand / the price you paid for this reprieve:
how it will change you / your face, your health, your self.
be glad, / they will tell you, averting their eyes
you will find out / that it was too high.
open your eyes. / rise
destruction is all you can see / from your bed; it has become
a battlefield, / a prison.
bloody lips and broken skin / this body no longer yours
a corpse, speaking. we know you / will fail to achieve your goal.
we know you will try. / find another way.
dream. / plan
within the confines of your mind, / the rise and fall of empires
avenge with your blade / the death of thousands.
the ghosts all around you. / the next day, repent
close your eyes / start over.
this rage beneath your skin / the poison in your veins
the monster you’ve become / slowly kills you
it’s a promise. / you will keep it to yourself:
all this pain you’ve endured / it will be over soon
someday it will end / just as it began.
the spirits of the dead hold you up in their arms / the embraces of the living are nothing but
freezing wind blowing on your exposed skin. / the scattered ashes of broken promises.
for them / alone
you will stay / you will wait
you will live. / you will die.
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
✓ Live Streaming✓ Interactive Chat✓ Private Shows✓ HD Quality
Anya is LIVE right now
FREE
Free to watch • No registration required • HD streaming
“Why don’t you use ai” idk man beyond the obvious environmental and “this machine causes psychosis and encourages people to kill themselves” thing I think asking the equivalent of a solid D student who is also a pathological liar if they can answer my question/do the work for me seems pretty fucking stupid
Text of tweet under the cut because it is loooong.
But... Stochastic Parrots.
Timnit Gebru was fired from Google in December 2020 for refusing to retract a research paper, and every single warning that paper made about large language models has now happened at a scale the industry spent 4 years trying to make people forget about.
Her name is Timnit Gebru.
She co-led the Ethical AI team at Google. She co-wrote a paper called "On the Dangers of Stochastic Parrots" with Emily Bender at the University of Washington and two other researchers. The paper was 14 pages long. It was submitted to a top AI ethics conference. And it was the reason Google decided that one of the most senior Black women in AI research could no longer work there.
The story Google told publicly was that she resigned. The story she told, confirmed by 2,695 of her colleagues in an open letter, was that she was fired by email while on vacation because she refused to either retract the paper or remove her name from it.
The paper had not even been published yet.
Here is what she actually wrote, and why every prediction inside it has now come true.
The first warning was about scale itself. Bender and Gebru argued that training ever-larger models on ever-larger scrapes of the internet would produce systems that appeared fluent but had no actual understanding of language. They called these systems stochastic parrots because they would repeat patterns from training data with statistical confidence and zero comprehension. The paper predicted that this apparent intelligence would fool both users and developers into trusting outputs that were structurally incapable of being reliable.
This was 2020. GPT-3 had just come out. The paper predicted the hallucination problem before anyone had a word for it.
The second warning was about bias amplification. The paper documented in detail that internet-scale training data contains systematic overrepresentation of dominant viewpoints and underrepresentation of marginalized ones. The models would not just absorb this bias. They would amplify it, because the optimization process rewards confident outputs, and confidence in language patterns tracks frequency in the training set.
The prediction was that hiring tools built on these models would discriminate against women. That healthcare triage tools would underperform on Black patients. That loan approval systems would entrench inequality while presenting their decisions as neutral algorithmic judgment.
Every one of those things has now been documented in deployment.
Amazon's hiring algorithm penalized resumes that contained the word "women" in any context. Healthcare risk scoring algorithms used by major US hospitals were found to systematically underestimate the medical needs of Black patients. Apple Card's credit algorithm gave wives credit lines 10x lower than their husbands for the same financial profile.
The third warning was about environmental cost. The paper calculated that training a single large language model produced emissions equivalent to the lifetime output of 5 cars. The prediction was that the race to scale would create an environmental footprint that would eventually rival entire industries.
In 2024, Google's emissions were up 48% from 2019, and the company explicitly blamed AI infrastructure. Microsoft's were up 29%, same reason. Both companies have now quietly abandoned the climate commitments they were publicly celebrating the year Gebru was fired.
The fourth warning was about documentation. The paper argued that the training datasets being assembled were too large for anyone to actually audit. Nobody at Google, OpenAI, Meta, or any other lab could tell you with confidence what was in the data their models were trained on. This was not a temporary problem to be solved later. It was a permanent feature of the approach.
In 2023, researchers discovered that the LAION-5B dataset, used to train Stable Diffusion and other major image models, contained thousands of images of child sexual abuse material. The companies that had trained on the dataset had no way of knowing. The paper predicted that category of failure 3 years before it was found.
The fifth warning was the one Google cared about most.
Bender and Gebru argued that the deployment of these systems would centralize linguistic and cultural power in the hands of the small number of companies that could afford to train them. The internet would become a place where the dominant voice was a statistical average of dominant voices, presented as a neutral assistant. Languages underrepresented in the training data would degrade over time as more web content was generated by these systems and fed back into the next training run.
This is now happening in real time. A 2024 study found that 57% of new web content in English is AI-generated or AI-assisted. Researchers studying low-resource languages have documented active degradation in translation quality, because the synthetic content fed back into training is itself worse in those languages.
The paper Google fired her for predicted the model collapse problem before model collapse had a name.
The mechanism behind why this all happened is the part of her work that nobody quotes.
Gebru's argument was not that AI is dangerous in some abstract sci-fi sense. Her argument was that AI is dangerous in a very specific structural sense. The technology was being built by a small group of researchers who shared similar backgrounds, worked at similar companies, and were rewarded for shipping products faster than competitors. The incentive structure made it impossible for safety, ethics, and bias concerns to slow anything down. Anyone inside the system who raised those concerns was either ignored, sidelined, or removed.
She was making that argument from inside Google.
Then Google proved her right by removing her.
The team Google had built to make sure their AI was safe was dismantled in 90 days because they did the job they had been hired to do. Margaret Mitchell, the other co-lead of the Ethical AI team, was fired two months after Gebru for searching through her own emails for evidence of how Gebru had been treated.
Gebru did not stop. She founded DAIR, the Distributed AI Research Institute, in 2021. The mission is to do AI research outside the control of the companies that have a financial interest in not hearing the answers.
Every prediction in the Stochastic Parrots paper has now been validated by deployment. Hallucinations are an industry-wide problem the largest labs cannot solve. Bias amplification has been documented in hiring, healthcare, lending, and criminal justice. Environmental costs are larger than entire small countries. Training data audits remain impossible. Model collapse is an active research crisis at every major lab.
The question worth sitting with is the one almost no one in the industry will say out loud.
Every researcher with the technical credibility to call out these problems watched what happened to her in December 2020 and made a calculation about their own career. The number of people willing to speak publicly about safety and ethics issues inside the major AI labs collapsed after that firing and has not recovered.
The researcher Google fired for warning about exactly what is now happening was right.
The company that fired her is now the second-largest deployer of the technology she warned about.
And the people inside that company who agree with her are not allowed to say so.
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
✓ Live Streaming✓ Interactive Chat✓ Private Shows✓ HD Quality
Anya is LIVE right now
FREE
Free to watch • No registration required • HD streaming
You ever think about many peices of media have zero women and thats just perfectly normal but if a peice of media has an all female cast people get... like that? Women should be allowed to kill over this btw
At 1 PM on a Friday I get an email from my boss. I'm busy as hell so I don't check it immediately. Then I get a phone call from my boss, which has almost never happened before. I'm a white collar worker, a historian. There's never a 'historical emergency' requiring a phone call to kick me in the ass and get to work.
The request is so urgent my boss needs it by the end of the work week. Which, y'know, is 5 PM on a Friday. So I have four hours to do it.
It's a forwarded request. Somebody contacted a member of the donation team asking for help, "I need a map from the Vietnam War to use for a presentation." It's somebody she's trying to coax into giving a five figure donation to the museum.
The request was asked to the donation team member, who then emailed my boss, who then emailed and called me urgently.
This map required:
North and South Vietnam in it
All four areas that South Vietnam was divided into for military purposes ('Corps') clearly delineated
Four cities, all of them horrifically misspelled, and only identifiable because I know what battle the requester is asking about (it’s in III Corps on the border with Cambodia) (the requester danced around the battle but I’m knowledgeable enough to identify it)
Has Laos and Cambodia in it
Has the Ho Chi Minh Trail in it
So. I was mad about the 'you have literally four hours to find a map with a lot of requirements.'
I was then mad at myself about finding a copyright free map from Texas Tech University within half an hour, proving her right for asking me to do it.
Then, after I found a map that perfectly met the requirements, I was equally amazed, baffled, and horrified when I read further into the forwarded email chain.
The donation team team member they were speaking to used AI to generate a map.
The above put half of North Vietnam in South Vietnam, made the Ho Chi Minh Trail a country, made 60% of Cambodia part of South Vietnam, put the DMZ extremely high up in North Vietnam, completely disconnected the southern tip of Vietnam, misplaced all of the Corps zones, etc etc
At the very last second the donation team member had a moment of divine clarity, remembering there's three historians on payroll to ask for this kind of thing from. So she contacted my boss while saying, "I had fun with this, but I decided I should check for accuracy before I send it to the donor! I need a fact check by the end of the day, then I send it"
My boss, while not the most knowledgeable on the Vietnam War, does know her geography. She took one look, and knew it was so off she called me to tell me how urgent it is that I look at the email and respond
good fucking god, jesus tap dancing goddamn christ, I'm glad I was asked to look at it and then find a real map
My fear has never been that AI would replace human intelligence. My fear has been that the people who Know Things and the people who Make The Decisions are almost never the same people.
We’re throwing real intelligence out on the street to starve while worshipping the shambling Frankenstein-ed corpse of knowledge puppeteered by those who see us as disposable assets.
Just watched Adam Conover (of Adam Ruins Everything) make such a solid point that I think we should spread far and wide. Yes, having AI write your emails is lazy, sure, but people love being lazy. We need to really emphasize that sending AI emails (or using AI responses on social media, or publishing AI flyers, or or or) is rude.
It's rude. You're making someone take their time to read something you couldn't bother to write. You're telling them they were so unimportant you couldn't be bothered to actually take the time to say something yourself. And frankly, you're lying about it while you're at it.
The above is doubly true if the content of the email is something that will be important to the person receiving - especially something that affects them negatively. They see that this thing that affected them so much didn't matter enough to you to write it yourself. I was a bystander to such a thing not long ago and it was just awful.
If I may offer the lecturer's perspective on this idea:
Currently, it's marking season for us in the UK. I have an exam board in four hours, in fact, which is where we all go over every profile of every student on our courses, see what results they've achieved, and work out their "decision" - if all is well, the decision is to let them continue the course, or the final degree grade calculated if they're in final year. If it hasn't gone well, the decision is about whether they get to rework the pieces that failed, resit exams, repeat the whole year, or be required to withdraw.
And, as has been the case for the last two years, the profiles are now littered with plagiarism investigations. Every one of those - every single one - will have come in as an assignment that the lecturer received, and started reading, and then with a sinking feeling thought "This isn't your work." Every one had to go to an academic misconduct hearing. Every one is an enormous draw on time and resources, including the emotional reserves of the lecturer.
And I know that's not the main issue! I know in the grand scheme of things, our feelings aren't the most important part of this equation! But as we're talking about rudeness, let me explain:
Firstly, the work itself. You begin reading, you see it's AI. Contractually, we have to read it anyway, and give feedback on why it's shit, and what makes it bad, and that is absolutely fucking soul destroying. Most students who use AI are doing so because they've managed to train their brains to find reading something boring abhorrent, and they want to skip that part; but a ChatGPT-generated report is bland, vague, and utterly devoid of any passion, insight or personality. In short, it's boring. You simply passed your boredom on to us.
Secondly, regardless of your personal feelings about the assignment, it at least had a purpose. It was there to stretch you, and make you think about the topic so you could learn about it, and to test that learning so we can all make sure you have actually learned what you need to. But the slop you handed in, that I now have to mark? What's the point? Literally what is the fucking point of me marking it? You didn't even write it. None of the feedback I'm obligated to give means anything to you. I'm marking ChatGPT, and it can't read.
Which means, not only is it fucking boring, it's actively pointless. Ask anyone in the world what a boring but pointless obligatory task does to your mood. Imagine that.
Thirdly, the misconduct hearing. Because listen, again, the lecturer's feelings here are, once again, not the main point. Students who cheat like this aren't doing so because life is hunky dory. They're stressed and overwhelmed and struggling, and they think they've found a magic way out, and so being pulled into a misconduct hearing - where the best they can hope for is to have to redo the whole piece for a capped mark, on top of all the rest of the work they have (functionally, a bonus assignment), and the worst is expulsion - is a mental breakdown-inducing experience. That, obviously, is the biggest issue.
But, the lecturers know all that, which means we know what we're triggering if we do report it. I cannot tell you how upsetting it is to receive a slop assignment, realise what it is, and then have to make the call to report it. I know damn well how upsetting that's going to be for you. I know how stressful and painful that's going to be. I know this might mean you're going to be thrown out of university. In some cases, I know it means you will be.
I know I could look the other way to spare you that
And oh, that gets tempting. When things are really bad for you, and I see you struggling, and this is your third strike; fuck me but it's tempting to pretend that I can't tell.
I cannot do that.
Which brings me to number four: the soul-bleachingly fucking horrible ordeal that is the misconduct hearing itself. Most people are non-confrontational; I'm no exception. I also simply do not enjoy a sobbing, panicking student sitting in front of me, telling me about how stressed and scared they are and how they're terrified they're going to fail. But that's how these things go.
Our most recent example is an international Masters student. I don't know the particulars for him; but I do know it's not uncommon in his part of the world for families to go into obscene debt, often to loan sharks, to send their kids to UK universities. Failure means more than just academia for him. Having to sit through him turning white and quietly begging us to give him another chance before he left in tears he tried to hide from us was, obviously, much worse for him than us; but it was honestly traumatic. Even now, two weeks later, I can't get it out of my head. There's nothing we can do; but, I feel guilty anyway. I could have looked the other way.
(It wouldn't have passed anyway. It was terrible. But at least he'd probably be allowed a resit - we're still waiting on the outcome of this one, but he may well be withdrawn)
To bring this back to the point of the post:
I know my feelings aren't really the ones that matter here. I do know that. But, every time a student chooses to use AI to write an assignment, all that is what happens behind the scenes. My job nosedives into being shit. Whether it's reading the boring slop, having to write pointless feedback, or making the upsetting decisions to report it when I know what the consequences will be and then having to deal with the guilt, my job that I love suddenly becomes shit. And that, actually, among the many other things it is, is fucking rude.
I don't know who my intended audience is here, so whoever needs to hear this, I am begging you to learn to participate in conversations that are about things you aren't interested in.
Part of socializing and having friends is being a good listener even when you don't actually give a shit about the subject.
Your are hurting other people's feelings when you bluntly respond with "Anyway..." and then change the topic.
It can not always be about your preferred topic.
You are being rude. Yes, even if you are neurodivergent. You can be both autistic and rude.
Autistic mom of two autistic kiddos here. I do something with my kids called playing the good friend game. I have them ask me a question like “what did you do this weekend” and then I respond with something they don’t like ie: “I ate some sushi for dinner on Saturday! It was delicious!” Their job is to not make a gross face and not to yuck my yum. They have to ask me a follow up question! Minimum one about what I liked about the sushi or who I went with or what I ordered etc. and they have to be kind. I change up what I answer with for the initial question and for all follow ups. It’s incredible!
You can learn how to chit chat about things because you care about the people around you. Small talk is banal and boring but the people in your life are real and lovely and nuanced. Get to know them. You can do it. You’ll build relationships that last. Don’t do it for the sake of small talk. Do it because you’re deeply invested in learning how to see the people around you as whole pictures.
To be seen is to be know. And to be known is to be loved. Love the people around you and that love with come back to you.
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
✓ Live Streaming✓ Interactive Chat✓ Private Shows✓ HD Quality
Anya is LIVE right now
FREE
Free to watch • No registration required • HD streaming
Pre-manufacturing cultures will really be like, here is the most elegant and gorgeous outfit you can imagine, and it’s achieved entirely with rectangles, ropes, and pins.
Over and over again, across the world, in cultures that never even knew of each others’ existences. Just, rectangles, knots, and pins. And I love that for them.
If we wanted to engage in nuance (lol, lmao) on the "are audiobooks reading" debate, we really do need to bring literacy, and especially blind literacy, into the conversation.
Because, yes, listening to a story and reading a story use mostly the same parts of the brain. Yes, listening to the audiobook counts as "having read" a book. Yes, oral storytelling has a long, glorious tradition and many cultures maintained their histories through oral history or oral + art history, having never developed a true written language, and their oral stories and histories are just as valid and rich as written literature.
We still can't call listening in the absence of reading "literacy."
The term literacy needs to stay restricted to the written word, to the ability to access and engage with written texts, because we need to be able to talk about illiteracy. We need to be able to identify when a society is failing to teach children to read, and if we start saying that listening to stories is literacy, we lose the ability to describe those systemic failures.
Blind folks have been knee-deep in this debate for a long time. Schools struggle to provide resources to teach students Braille and enforcing the teaching of Braille to low-vision and blind children is a constant uphill battle. A school tried to argue that one girl didn't need to learn Braille because she could read 96-point font. Go check what that is. The new prevalence of audiobooks and TTS is a huge threat to Braille literacy because it provides institutions with another excuse to not provide Braille education or Braille texts.
That matters. Braille-literate blind and low-vision people have a 90% employment rate. For those who don't know Braille, it's 30%. Braille literacy is linked to higher academic success in all fields.
Moving outside the world of Braille, literacy of any kind matters. Being able to read text has a massive impact on a person's ability to access information, education, and employment. Being able to talk about the inability to read text matters, because that's how we're able to hold systems accountable.
So, yes, audiobooks should count as reading. But, no, they should not count as literacy.
It sucks being a philosopher of mind and language when people who are not linguists or cognitive scientists start reblogging critiques of three-cueing. Not because I think it's a good method for most reading environments or because I think there is any good evidence to support it (I don't), but because every single goddamn post critiquing it also leaps to wild conclusions about other things to do with reading, language acquisition, and the mind in general.
Just a few things of note, off the top of my head:
Yes, you are a pattern-recogniser and that is how you read the vast majority of the time.
No, you do not read every single word precisely - there are acres of studies about this, tracking eye seccades as people read. Your eyes jump about on the page, they do not smoothly scan from one side to another. This is a fact. It was not made up to support three-cueing.
Yes, it's how LLMs work. LLMs were designed to mirror (on a very simplistic level) our brain processes and language learning processes. It's not an accident or an indictment; the LLMs are based on US.
It does not follow from this that LLMs understand what they receive as input - this requires more than pattern detection. Arguably, LLMs recognise nothing, they are just trained to spit out certain patterns in response to other patterns.
It also doesn't follow from the pattern recognition methods of experienced readers that we should teach new readers to read by three-cueing. This is something experienced readers do because they already know an incredibly vast number of words and this enables them to recognise the rough shapes of words at a glance - and is also why even professional proofreaders miss things sometimes, because they see roughly the right pattern of letters they expect and not necessarily the precise arrangement the symbols have on the actual page.
Bad reading comprehension is not new or limited to people who learned to read using three-cueing, as evidenced by all the people writing all these rants who appear to have missed the nuance in the words of the experts in the articles they themselves are linking to.
The biases you bring to reading a thing are much more likely to prejudice your understanding than the way you learned to read, even if it's also true that the way you learned to read did you no favours.
In short: you, too, are capable of pissing on the poor.
to anyone in the areas impacted by the wildfire smoke, my #1 biggest piece of advice as someone whos been dealing with wildfire smoke in the NW united states for years, is build yourself a Corsi-Rosenthal Cube
they perform as well as expensive HEPA air cleaners, and are comparatively VERY inexpensive. all you need is a box fan, 4 air filters, a piece of cardboard, and some duct tape!!!!
i think it took us maybe a half hour to put ours together, if that, and we replace the filters every 3 months. it's really made a HUGE difference, both when the air quality is bad, but also with our allergies
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
✓ Live Streaming✓ Interactive Chat✓ Private Shows✓ HD Quality
Anya is LIVE right now
FREE
Free to watch • No registration required • HD streaming
my close friend from uni was a cis girl who had the audacity to wear pants and cut her hair short and like nobody at this school, a place OBSESSED with ‘respecting everyone’s gender identities,’ would call her ‘she.’ after MONTHS of this she started wearing a fucking pronoun pin to work and i dont even think that fixed it. me, im sorta androgynous; i have shaggy self-cut hair and go by a neutral name, but i always say my pronouns are she/her, and people ive worked with for months and have introduced myself in front of fifty times will STILL reflexively say ‘they’ for me. i respect the progressive circles i run in, but this IS evidence of misogyny. people’s definition of “woman” or “girl” is so narrow and high-maintenance that even the tiniest deviation from the norm gets you forcibly defeminized. but it’s a compliment, right? like who would wanna be a girl anyway?