Missing in Action

Mierel he/him they/them, 19 (probably won't be posting any art for who knows how long)

[miereldraws]

Hello random user I am Mierel (pronounced My-rell) and I'm an aspiring artist and game designer. Here are some things to know about me:

I will sometimes draw stuff and post it but for now it will probably be mostly reblogs.

I will never hesitate to fixate on one specific thing and ignore anything else.

I'm asexual and demibiromantic.

DNI IF YOU ARE:

LGBTQ+ PHOBIC, A MAP, A TERF, RADFEM, RACIST, A GENERALLY MEAN PERSON

[leighlew3]

ALL. OF. THIS.

[tieflingkisser]

[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.

[vortexsophia-deactivated2022110]

[veerletakino]

[no-one-offical]

Reblogging thrice for effective reach

[homophobla]

[homophobla]

Today is litteraly the only day in all of human history where it is physically possible to reblog this. Given that I posted it today

[wifegideonnav]

“bits to use in everyday conversations”

[teaboot]

Wild that folks keep saying beekeepers abuse bees as if bees are not both venomous flying animals and fully unionized

[teaboot]

Hubris to think you COULD abuse bees

[teaboot]

I think that if you see a balrog in an unlocked kitty kennel then you can assume that it wants to be there

[vegance]

Actually, beekeepers take many precautions to keep their bees from leaving.

many clip the wings of the queen, destroy new queen cells, cull queens they don't like and use bee pheromones to prevent a hive from naturally swarming or absconding. They also try and prevent mating with the African honey bee, which makes them less docile among other things. During artificial insemination of queens, drones are crushed and „spare“ queens are killed.

and commercial beekeepers even cull their hives during winter, or when they are not producing well.

Coupled with the fact that there is evidence that insects do feel pain, this is not great.

(Not to mention that honeybees are an invasive species in most places, competing with native pollinators and spreading disease)

[tetrafelino]

It is so fucking bold of you to link the exact same paywalled book thirteen times in your reblog to throw people off the fact that you're using one source from 1859.

[champawattigress]

I don't think there's ever been a funnier vegan response in the history of this hell site. This is actual gold. If you cited an argument like this in a first year undergrad intro to bio module, then your lecturer would be legally entitled to fire you out of a cannon. I think I'm dying

[lost-where-the-forest-would-grow]

I just got described as an "ad hating commie" by someone because I said a minute of youtube ads is unpleasant. fully spent 5 minutes arguing and defending youtube ads. insane stuff

[daovihi]

reblog if you are an ad hating commie

[cosmonaut-observer]

If you live in the UK you need to see this

Protect Internet Freedom from now until forever. It's important existentially! Americans stand with UK citizens in our struggle against government censorship

[russettrush]

Growing up in the online world: a national consultation

We are consulting on further measures to prepare children for the future in an age of rapid technological change. This includes potential ag

Got the link via @finalducc

If you live in the UK, please be sure to take part in this!

[fecklesheckleshacklesschmeckles]

closes 26th may 2026!

[bluerose5]

Thank you, Black people in fandom spaces. Thank you, Black creators and Black lurkers. Thank you Black artists, Black writers. Thank you, Black bloggers, Black influencers. Shoutout to those Black characters, both canon and original. Thank you, Black people, both queer and cishet.

Your perspectives matter. Your representation matters. You are not bothersome for demanding equal treatment in fandom. It is not your responsibility to make fandom more welcoming and inclusive to you. It is not your sole responsibility to create all of the Black-centered content. You are not "ruining" anyone's fun for demanding better for yourself, and anyone who says otherwise can go fuck themselves. Any fandom worth being a part of should have no room for racism in it.

Black people in fandom, you are wanted. You are needed. You are loved and appreciated. Don't let anyone tell you otherwise.

And since they don't get told it near enough, thank you, Black women especially!!!

[murfpersonalblog]

You are not "ruining" anyone's fun for demanding better for yourself, and anyone who says otherwise can go fuck themselves. Any fandom worth being a part of should have no room for racism in it.

[gahdamnpunk]

Can we support him please?!

[comfygiant]

I would love to share this with everyone who may happen to see this post. Please support this wonderful human being. He spent nearly a half century in prison for a crime he never committed. And the only thing that kept him going was his artistic endeavors. He deserves the best life can offer anyone ❤️

[mildmayfoxe]

HERE IS A LINK TO HIS WORK

[destielmemenews]

source 1

source 2

[kaban-bang]

[kimmycup]

Like I don't think you guys comprehend what happened in Poland just now but everyone needs to be talking about it.

A random influencer decided he'll listen to an anti cancer song on loop. People liked it enough times he ended up listening for 9 days.

He raised 90 million in these 9 days, and then 160 million more over the last 10 hours, for a total of 250 million.

Hundreds and thousands of people signed up to donate marrow.

Hundreds of celebrities shaved their heads in solidarity.

The Foundation receiving this money had to create a special commission to figure out how to distribute the money.

The national TV stations got highjacked to stream this for hours because it was better news than anything happening in the world.

Because we broke and DOUBLED the world record for this kind of thing.

They raised about as much as the biggest running charity event in Poland did in a whole year with three decades of tradition and a goddamn army of people.

And they did it on a goddamn amateur set up in a shabby room sitting on folding chairs.

[kimmycup]

Little update: people keep donating despite the stream being over. We're at 280 million in the fight against cancer.

Łatwogang refuses to collab with companies that only reached out to him now because of popularity or give interviews. He said any medals people wanna give them should go to the doctors and nurses and the cancer patients.

Someone offered to renovate that shabby little flat for him as a thank you. He refused.

Someone counted up how many people appeared in that room during the whole thing - it was 319 total.

[belladonna-rx]

Love to hear it. Love news articles too.

https://www.reuters.com/business/healthcare-pharmaceuticals/polish-nine-day-charity-stream-breaks-records-support-kids-with-cancer-2026-04-27/

Just love this headline phrasing.

Streamer crushes MrBeast charity record with $70 million raised for kids with cancer - Dexerto

Polish streamer Łatwogang’s charity livestream has raised over 203 million złoty for children with cancer, breaking the charity stream recor

[paper-mario-wiki]

this is really frickin neato

[stinkyhat-deactivated20231213]

https://store.steampowered.com/app/948220/ColorTool/

it’s coming out tomorrow!

ColorTool on Steam

Create palettes by placing an intricate web of colors and their connections. See directly how the palettes change the look of your illustrat

[septicake]

pretty sure a few artists follow me for whatever reason so yeah

[official-lucifers-child]

[gay-for-frog]

Rapists, and killers, too? Really? (Those on death-row?) The drug/prostitution problems are just a portion of USA criminals.

[official-lucifers-child]

yes, all criminals. the moment you say “except X criminal” is the moment that people will try to convict their opponents as having committed X crime.

it’s the same thing as what’s going on right now with people equating drag to some sort of child exploitation. “but the children!!” they wail, and people listen because oh, if drag is harming children, then drag MUST be BAD, so we HAVE TO BAN DRAG.

do you understand what i’m saying? you can’t take away the rights of any category of criminal, because suddenly that category will be overflowing with people who totally 100% definitely committed that horrible crime.

[captain-acab]

Just to look at this from one step further back, let's accept the premise. Let's imagine that there is a type of crime that is 1) horrible evil irredeemable universally-agreed-upon bad, and 2) could somehow be prosecuted with 100% accuracy and 0% bias. Yes, even then, those criminals deserve a right to vote.

Do you they're going to like vote themselves out of jail? Vote to make murder legal? What exactly are you afraid of.

Realistically they'll just... vote just like anyone else. They'll help elect city councilmembers they think will better their hometowns, and presidents they think will best serve the country. They might even vote in their own interests! To reform prisons, fund rehabilitation programs, and outlaw predatory practices by telecoms. Are you saying you don't want any of those things?

And even if there were one of those super-duper-unambiguously-evil totally misanthropic death-row convicts, who's scheduled to be execute the very next day and just wants to sow chaos and watch the outside world burn however they can... what's the worst they could do, vote republican?

Taking people's rights away isn't bad because it might happen to someone you like, it's because taking people's rights away is bad.

[heartspocky]

do people think that convicted serial killers are going to vote to legalize serial killing

[primrosetta]

me: Princess Zelda’s a pretty cool character and is way more than just a damsel in distress.

some nerd: source????

me: 

[primrosetta]

me: ??????

me:

[se-meme-drakant-deactivated2025]

Didn’t do shit in Link’s awakening though

[primrosetta]

This post has over 50 thousand notes and this is the most baffling response to it I have ever seen. 

Yes, Zelda didn’t do anything in a game that she was not in. You fucking got me. In other news, Mario is a bad character because he didn’t do shit in Metroid Prime. 

[primrosetta]