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.
Like they tried to change Reblogs and people rightfully got up in arms, this is a LOT worse. In order to have access to any sort of thing dubbed mature, and We haveALL seen what they think is mature, Everything from a black and white photo of a black woman's arm, to posts about IUD recalls, to a nude painted by a 17th century artist, to anything involving the word Trans; you have to send your personal information to a third party site that WILL get hacked, and you will be doxxed. And they can say "Oh shit, well it wasn't us who sent your name address and gender identity to Moldovan teenagers, here's a couple extra minutes in the ball pit.
That's bad enough!!!!!!!! But the entire idea of needing permission from state authorities to access anything labeled mature by our friendly AI overlords is some fucking Boll shit. Die Gedenken Sind Frie baby. This is all a reaction to people getting uppity about their lowly lowly rights and is being propped up by the same bad actors tht have made life unlivable. Fuck that shit.
"Well it's only being rolled out in Brazil and UK" Yeah, to start. "Well they're being forced to do this by laws." YOu know it's always really funny when these tech giants (Or whatever you call owning tumblr dot com) get really antsy about laws considering they pick and choose which ones they abide by.
This is a breaking point and it's going to be very interesting to see how we proceed from here.
Tumblr is a place to express yourself, discover yourself, and bond over the stuff you love. It's where your interests connect you with your
select feedback from the category drop down menu, and tell them how HORRIFIC of an idea they age verification crap is!!! (including the reasons stated in the posts above)
also go to the @staff and @tumblr blogs and leave as many comments as you feasibly can about how awful this is.
REBLOG THIS SO MORE PEOPLE SEE IT!!!
IF WE GOT THEM TO BACK DOWN ON THE REBLOG CHANGE WE CAN PUSH AGAINST THIS TOO
(original tags: she literally had tears in her eyes while she explained that her pikachu had a boy tail shape. i was like hey she can still be a girl. all we know from her body shape is what her body is shaped like. maybe if we asked her she would say that she feels like she’s really a girl. and my kid was like BUT POKEMON CANT TALK 😭😭😭😭 so i was like ‘ok. pikachu. if you want to use he/him say ‘pika’. if you want to use she/her say ‘chu’. and if you want smth else say ‘pikachu’’. and my daughter had her say ‘chu’ and i was like see there you go! now. she doesn’t HAVE to change her tail shape. she can totally be a girl with a rectangle tail. but if she would feel more comfortable with heart shape bc she prefers it, or so that people won’t assume she’s a boy, i can make that happen. so pikachu what do you think? do you want tail surgery? and pikachu agreed enthusiastically! so. plushie gender-affirming surgery first thing tomorrow i guess!)
her results look great, congrats pikachu! 💖⚡️🏳️⚧️
just a reminder that this blog is run by someone who:
— is anti ICE & fascism
— is pro-choice & feminist
— supports trans & queer people
— hates generative AI & capitalism
— supports immigrants & people of color
— is pro-environmentalism & social justice
— supports palestine & all other territories unjustly suffering
I'm Vivvie, a neurodivergent trans woman who keeps getting trapped in retail jobs.… Vivian Oriente needs your support for Support Vivvie's M
hey, so i'm going on an unpaid leave of absence for a few months here soon to try to get over my burnout and catch up with life. i set up a gofundme to help with financial obligations during that time.
Political pet peeve but nothing being done is a "distraction" from anything else. Everything that the current administration is doing is an extension of American imperialism and conservative American values. Iran is not being bombed to "distract from the Epstein files", trans legislation is not being passed to "distract from Palestine", none of this is happening just to "distract" you. And dismissing certain issues as "distractions" is a horrific excuse of them.
It is, however, happening all at once to overwhelm you. And there is a HUGE difference.
it's actually sad how tumblr killed fandom culture on their own website just like that. yeah, people can still post their gifs/art/fics, but making every reblog its own chain + essentially giving credit to rebloggers individually and making it nearly impossible to access comments/tags as the original poster entirely defeats the purpose of being a creator on here. and while i appreciate the posts already going around encouraging people to engage with op's posts directly rather than through a reblog, it's just going to be ignored like every post pleading with people to reblog and engage with content to begin with
PLEASE FOR THE LOVE OF GOD IF YOU HATE THE NEW UPDATE REBLOG THIS POST
[ PT; please for the love of god if you hate the new update reblog this post ]
I am organizing a lights out protest on tumblr, from March 20th 6AM UTC until March 21st 6AM UTC. It is best if as many tumblr users as possible can join this protest, as a mass downtime in users is the only way the tumblr staff will listen to us.
If you cherish this hellsite, participate. Do your bit. Every person counts.
Thank you for reading, and to @staff @changes: give us our tumblr back, or the people will migrate somewhere else. This is a threat.
ramadan mubarak! may this month be full of blessings and peace across the globe. and may we take the time to reflect on those still fighting to make this month one of peace
stormborns
izzy806
ellievsbear
annaxmalina
jbbartram-illu
dededos
tangentwithtanner
