andromeda, for example.

probably too old to be here. writes like she breathes. tries too hard.

[tyrantisterror]

I've watched a lot of dinosaur documentaries, and a lot of them include bits of narration talking about how the end of the age of the dinosaurs involved years upon years of ash from volcanic eruptions and the asteroid impact filling the sky. And I gotta say, dang, I feel bad those poor dinosaurs having to deal with days like today for years on end.

The documentaries always talk about them starving but honestly I wonder how many of them got taken out by lung cancer.

[snarling-through-our-smiles]

[hypokeimena]

image description: tweet by olayemi olurin @/msolurin, reading: There's no high like making a successful hail mary argument. A prosecutor once asked for bail on my client because he had like 63 convictions and I was like judge... 63 convictions sounds like a man who showed up to court at least 63 times. And the judge was like yea you right /end description

[staticbluue]

“People should pass a test before being allowed to have kids.” “Isn’t it scary how white people have this inborn capacity for evil?” “I’ll never pass because males and females have different skull shapes.” “Autistic people have a stronger sense of justice than anyone else.” “I don’t want AMABs in my space because they’re dangerous.” “You shouldn’t have access to hormones if you dress like THAT.” “Anyone who does something that awful isn’t human.” “Some people really shouldn’t be allowed to vote.”

This is eugenics. This is phrenology. THIS IS NAZI SHIT, YOU ARE A LEFTIST BUYING INTO NAZI SHIT. YOU ARE NOT IMMUNE TO NAZI SHIT.

[skies-diary]

I've seen a lot of people in the comments saying people needing to pass a test to have kids is a good idea, but I don't think they understand what that means. The US government used to take Indigenous children from their parents for reasons as asinine as grandparents living in the home (called "overcrowded environment") or using only their native language at home instead of English (because it "failed to prepare" children for a world where English is the dominant langauge).

Children placed in residental schools were abused, murdered, and sold into arranged marriages with white men. More recently, children "adopted" out into white households are often also subjected to the same, as well as forcibly converted to Christianity.

The most effective way to protect children in the country is to give them rights to ensure their safety, autonomy and dignity. Eugenics and laws to control reproduction will not do that. Crafting legislation with the help of child wellness experts is a better choice.

[skies-diary]

I should add - the US government STILL disproportionately targets Indigenous families for child welfare checks. ICWA is supposed to protect against this but is unevenly enforced, and there's even talk of overturning it. Protect ICWA and Indigenous children.

[microsff]

The parking attendant paused by the double-length bay. Intended for mobile homes and cars with trailers, it was currently occupied by a sleeping dragon.

No parts of it extended beyond the lines, and the paper ticket was clearly displayed, impaled on a horn.

The parking attendant moved on.

[lionessfeather]

I was going to just queue it for later but then it stuck in my brain, and I decided to make it everyone's problem

[mini-wrants]

The thing that’s always missing from the “women didn’t fight for the right to work they were already working they fought to get paid” is that many women also very much wanted to work.

Women wanted to be lawyers and engineers and chemists. They wanted to use their brains in challenging and interesting ways. They wanted to get the satisfaction from solving problems and inventing new shit and getting attention for it.

I know not everyone is born with intellectual curiosity or drive or determination but some people are and many of those people are women.

[mini-wrants]

[mini-wrants]

Literally.

[thatdiabolicalfeminist]

"Dogwhistles" are called that based on real "dog whistles", many of which are pitched too high to be audible to most humans, but are still perfectly audible to dogs.

Rhetorical dogwhistles are things that most people would not be able to identify as bigoted because they're INTENDED to have plausible deniability.

The wording changes are meant to be subtle enough that people really can mix them up accidentally when they don't have much information!

So when the gender critical movement calls trans women "transwomen", they're hoping for 2 things.

this usage will spread enough among people who don't know any better, to give transphobes the plausible deniability of not LOOKING like a frothing bigot constantly.

the usage grammatically places trans women in a different category than other women; it's changing "trans" from an adjective to part of a noun to make this distinction.

The more we can avoid this usage, the less deniability have transmisogynists have when using it, and the less rhetorical ground we cede to the degendering and misgendering of trans women.

I don't actually think ceding this ground is LESS divisive than pointing it out politely.

[iguanamouth]

forbidden knowledge

[coldfruit123]

Can we see this frame by frame? Please and thank you.

[iguanamouth]

Yes Here Are All The Frames

[pibblepunk-deactivated20200219]

THAT IS NOT WHAT WAS ASKED FOR

[waterbottoe]

How dare you hide this in the notes

[toebeens]

wow that last comment really does deserve to be front and center, it’s perfect

[aph-shy]

@the-zebra-dragon

[horreurscopes]

never forget the universal rule of the order of things: People Will Not Read It

[horreurscopes]

signs at stores? émail? menu ?? instruction ? post online ? caption with andswer to question ? group hand outs ??? street sign ??? no. The Written Word Is The Enemy

[lasrina]

#The number of compliments i have gotten for reading a thing

The ability to occasionally Read A Thing will make you a hero in your workplace, especially if it is for example an error message that tells you what you need to do differently, or instructions on unjamming a printer.

[stephaniematurin]

how dare you say we put jam in the printer

[berkelbites]

Ok reblogging this again because story time.

I work in tech, and much of what I do is support sales reps within the company by resolving errors with the software they use.

There is one sales rep who, every single time I send her a message or email with extremely specific instructions that will resolve her issue, does something completely different from what I tell her. Every time. Without fail. It is so glaringly obvious that she has never read even a single word that I have written to her.

So one day, she sends me a message that says little more than "(software) is broken, help"

So I do my standard song and dance of asking her what she's trying to accomplish, and what specifically is stopping her from doing that. And eventually, after much unnecessary back and forth, she tells me there's an error message. I ask her to send me a screenshot of the error message. She does.

The error message basically says, "these two required fields are blank. To resolve this, please fill in these two specific fields, and then click save."

So I take a few deep breaths.

Then I lie to her.

I message her back, saying "hey yeah, for some reason it's not loading that screenshot on my end. Could you type out the full text of the error message for me?"

She does.

I ask her if she still needs help.

She does not respond.

[samueldays]

I have similar story from tech support.

Client is reporting that Some Thing Program doesn't work. I ask if there's an error message with further information about what's not working. Client says "no". I go over and ask Client to open Some Thing. Client double-clicks on the icon for Some Thing, it starts to boot, an error message dialog flashes up on screen, Client closes error message before I can read it, Thing closes after the error.

"What did that error message say?" I ask.

"What error message?" asks Client.

I tell Client to open the Some Thing again and then not click anything else. Client opens Some Thing, error message appears, Client clicks it away again.

I tell Client to stand up, step away, and give me physical control of the computer. I open Some Thing, start looking at the error message without closing it, and Client says "You should close that." I tell Client that I am reading the error message. Client is apparently accustomed to treating error messages as a kind of spam email that should be deleted as fast as possible, and gets agitated that I'm reading it.

I read the error message. It tells me what the problem is. I fix the problem. Some Thing works now.

---

Later, I start thinking about how such an error message might perhaps be engineered to be more attention-grabbing and close-resistant as a way of making people read it. It's not important for some random program here, but there are more important systems (medical, etc) where it would be reasonable to demand the user's attention because people's lives depend on paying attention to the error message.

But then people with a perverted intellect would still be thinking about ways to avoid reading the message, like dragging it off edge of screen or hiding it behind another window. So maybe the dialog box could have an always-in-front feature to override other windows, and the alert could use the computer's hardware "beep" functionality that can't be switched off by muting the regular sound system, and keep beeping... shit, I realize I'm reinventing pain, and get philosophical about it.

[doccyuk]

Story from The Past about My Mum:

She was a computer programmer / analyst, a... Long Time Ago. Called in for a system she'd installed before, the office folk said they kept having problems where it Didn't Work Right (no error, a malfunction)

She investigated, and told them that could only happen if they did 3 specific things in a specific order, which they should not ever do.

So, she asked, did they ever do that?

No! Of course not, was the answer.

So she made a couple of small changes, packed up and said that should be fine, but they should call her if there were problems.

The next week

She had a call saying "We're getting a strange error message on the system, can you help?"

She said, of course, can they tell her the error?

And the message was:

"You Said You Didn't Do This"

[liniini]

Katharine Hepburn in an interview with Barbara Walters (1981)

[boycritter]

"unbecoming" is such a great word. bro that shit was so rude you no longer Are

[whosname]

[id. A twitter post by @/Bennieeexyz Jury duty letter came addressed to my cat. Not a mistake. "Felix Martinez" - that's his full name according to his vet records. My last name. His first name. Somehow he's a registered voter now. Called the county clerk. Me: My cat got summoned for jury duty. Clerk: Is the name correct on the summons? Me: Yes, but he's a cat. Clerk: Is Felix Martinez a legal resident of this county? Me: He's a legal cat. Clerk: Sir, if the name matches our records, he needs to appear or file an exemption. Me: He can't file anything. He has paws. Clerk: You can file on his behalf. Me: Under what exemption? There's no box for "is a cat." Clerk: (pause) Check "unable to serve due to medical reasons." Me: What's the medical reason? Clerk: He's a cat. Me: That's not a medical condition. Clerk: It is if it prevents him from serving. Sent in the form. Got rejected two weeks later. "Insufficient documentation. Please provide medical professional's statement." Took the letter to my vet. Me: I need you to write that my cat can't do jury duty. Vet: Why is your cat summoned for jury duty? Me: Excellent question. No good answer. Vet: This is the weirdest request I've gotten. Me: Can you just write that he's medically unfit to serve? Vet: On what grounds? Me: He's a cat. Vet: (started typing) "Patient is unable to serve due to species-related limitations including inability to speak, read, or comprehend legal proceedings." Me: Perfect. Sent it in. Got another rejection. "Summons is mandatory. Failure to appear will result in contempt of court." My roommate thought this was hilarious. Roommate: Felix is going to jail. Me: This is serious. Roommate: Bring him to court. See what happens. Decided that was actually the only option left. Day of jury duty, put Felix in his carrier. Brought the entire paper trail of rejection letters. Checked in at the courthouse. Clerk: Name? Me: Felix Martinez. Clerk: (looked at the cat carrier) Is that Felix? Me: Yes. Clerk: (long stare) He's a cat. Me: I've been saying that for six weeks. Clerk: Why didn't you file an exemption? Me: I filed three. All rejected. Showed her the letters. She read through them, expression shifting from confusion to disbelief. Clerk: Someone rejected the veterinary documentation? Me: Twice. Clerk: (called her supervisor over) You need to see this. Supervisor read everything. Looked at Felix. Looked at me. Supervisor: How did a cat get registered to vote? Me: You tell me. Supervisor: This is a data error. Me: Took you six weeks to figure that out. They dismissed Felix immediately. Apologized for the inconvenience. Supervisor: We'll remove him from the voter registry. Me: Appreciate it. Supervisor: (pause) Out of curiosity, how would he have voted? Me: Probably whatever party supports universal treats. Got a formal apology letter a week later and a voter registration card. For me this time. Apparently I wasn't registered, but my cat was. Roommate: Felix committed voter fraud. Me: Felix committed nothing. He's innocent. Roommate: That's what they all say. Felix is sleeping on the jury summons now. Fitting end to his legal career. end id]

[thenerdyindividual]

The Tortall books are for horse girls, girls who wish they could talk with their pets, and girls who picked up as many chairs as their little bodies could lift when their teacher said “I need some strong boys to help me move these chairs”.

[biggest-gaudiest-poltergeist]

When we were children, my sister had private music lessons at her violin teacher’s house. I only visited there once, but I still remember that afternoon. The teacher had an artificial pond in her yard, a large beautiful thing with lily pads and plant life. And in the pond, there were goldfish. I had never seen such enormous goldfish. 

I spent several minutes just staring at them (and trying to convince them to bite my fingers.) When my sister’s violin lesson ended, her teacher came out to the yard and explained that these goldfish were the same small creatures that were often unfortunately sold in plastic bags at state fairs. They were only about two inches long apiece, when she bought them and put them in the new, empty pond. In essence, they were like every goldfish I had seen before, but they had been given a much larger, much richer environment in which to flourish. As a result, they had grown into some of the most remarkable, vibrant creatures my twelve-year-old self had ever met with. All because of a pond. 

Funny what lessons children remember. My sister doesn’t play the violin anymore, but that was the first time I caught a glimpse of the overwhelming extent to which it matters, the way the world treats us.

[yellow-hatco]

Reblogged again for this drawing I made for it

[221b-hound]

Give us room to grow and see how we flourish.

[justalurkr]

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.

[irispurpurea]

This is the paper. It's excellent, highly recommend reading it.

I remember reading about Gebru's firing but I had no idea this was the paper she was fired over.

[cheacegirl]

I think gatekeeping art is such a funny thing like. "Oh so this doodle I just made is art" yes next question. "So video games are art?" yes????? Bitch your essay you wrote on that book you didn't read in English class when you were 10 is art. Painting a fence is art. 2 sentence shitpost on Tumblr is art. Get on my level.

[bornt-urnge]

[postarcadian]

truly the idea of "this scene is (un-)necessary to the story" is such a fundamentally uncurious and anti-art way of engaging with fiction. the story itself is unnecessary, in the sense that all art is unnecessary, because art is not a fucking optimization problem. that's the beauty of it

[argumate]

“my bonnie lies over the ocean, my bonnie lies over the sea,”

are we talking about the same body of water here, which is weird, or different bodies of water, which is even weirder

[theinflammablemammal]

this makes more sense if you assume both are the same body of water but the time between the statements is about 50 million years

[argumate]

you’re right, that’s significantly less weird

[argumate]

toasthaste said: maybe there’s more than one bonnie

blocked

[lalaithion]

A hypothesized geography.

[truth-shitposting]

i bless the bonnie over africa

[dark-magician-girl-meets-world]

The bonnie is in geosychronous orbit, thus over all the Earth’s bodies of water

[special-lucky-chosen-one]

the bonnie is merely moving very fast, perhaps at hypersonic velocity, relative to the singer

[special-lucky-chosen-one]

my bonnie lies over the ocean

my bonnie’s exceeding mach 3