Inky Thinking

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.

ANIMALS:

catenarwhal: #mandatory 'how cows produce milk' mention#i'll never recover from that one I fear

piromantic: #one time i saw someone fake their way through describing how spiders behave

pluto-lichen: horses

misskittypotter: #stardew valley faking its way through what fresh fish smell like

pa-pa-plasma: #saw someone faking their way through knowing what a seal is once #i still am fucked up over that one to this day. they just straight up did not know #& they were NOT good at guessing it either like it was clear they had never googled that animal ever #& was only just now realizing via answering questions from anons that seals are not!! what they assumed. initially

SEX:

dykevandyke: #what a prostate is #and where it is located #as in. external.

dreamyeyedrose: #I remember back in the ff.net days reading an Ichigo/Renji fic where the writer assumed the penises go inside each other #and I was like “I mean I don't know how it works for sure I don't have one but idk if that's how it works”

SOME OTHER FOOD STUFF:

thetrekkiehasthephonebox: #add another one to the list bloggers#this character is cooking a salad

shosta: #still baffled about the published work that didn't know food could freeze

sun-dari: #once i read a fic where the author didn't understand cinnamon

alto-tenure: #read something recently where the author was just. blatantly wrong about spices

dramatic-dolphin: #i saw someone try to fake their way through what ramen is once. like 14 years ago.#but i remember.#i was very confused about ramen for a few months. they were writing it so authoritatively.

the-celery-stalks-at-midnight: #i will never ever forget someone putting leftover fries in the microwave to reheat them and setting the timer for five minutes

typeghost: #this sparked a memory of a hannibal fic where the author had to fake their way through writing about gravy

draculin: #the one fanfic where the author knows about coffee only as a concept wrote a character as a coffee drinker#was very interesting#I don't remember the fandom or the plot but I was mesmerized by the coffee actions and choices

11235811235811: #there's a lot of faking their way thru congee in the svsss fandom i'll also note

fishali3n: #read one where the person clearly didnt know what tofu is

emmy-everafter: #in the aftermath of shadow and bone s2 i saw a lot of people pretending to know what stroopwafels are #babes they are more like cookies than breakfast waffles #like yes there is a waffle pattern but you're not gonna cut into a stack of them with syrup and sugar#🤣🤣🤣

NON-FOOD STUFF:

red-umbrella-811: Shoutout to Dame Agatha Christie for faking her way through what a wrench is in a very popular published work.

bluebeetle: #once saw someone have a character put an entire phone book in their pocket

nonametis: #- sex talk in languages other than english #<- or just the petnames in a different language other than English

sadisticpony: #the fanfiction i saw this week where op DIDNT KNOW HOW AUTOMATIC DOORS WORKED #and that they arent in peoples homes!!! of course. also opening the automatic door for someone is unironically very funny but its not #its not like. grabbing the door handle to let someone in. helpppp

danmeichael: #reminds me of the fic with the figure drawing class where the character started with the feet. #i love you feet first figure drawing author

meowmix1100blr: #me watching this one fic absolutely obliterate what the board of directors does

vexedhexes: #one time i read an architect character making a doorway bigger by building a bigger door #what a beautiful world. #OH. also gravity falls fic where they go 'oh piedmont is in california so its warm all year round'

leveragehunters: #characters going to a beer garden #And it's literally a garden outside the pub#It was a very cute mistake

fitofpique: #yes! #grown men do not get blind drunk off two beers #but i am possibly guilty of the hypothermia one #assuming it does not make you very horny?

dadvans-likes: #always thinking abt the soup kitchen fic #the entire setting of the fic was 'soup kitchen' #and i very quickly realized #the author did not know what a soup kitchen was #and they thought that soup kitchens only served soup #fic

msmargaretmurry: #i love fanfiction #once read a fic where the characters played 20 questions #but the author seemed to not know how to play 20 questions and was just kind of winging it........ #immaculate

shakespeareaddict: #Look I know not all of us are hockey experts #But it takes about ten seconds of research or any attention paid to the show to realize #That the Stanley cup playoffs are not in fucking September

baejax-the-great: #the funniest one i saw #was someone faking what church is like #like 1. they really didn't have to write an entire church experience for their fic #and 2. they had clearly never even watched a show where people went to church #it was bonkers weird

twosunson: #things ive seen authors faking #knowing how to unclog a drain #knowing. literally any history #knowing what ketamine looks like (apparently- oregano) #(you know who you are)

waterhorseyblues-ao3: #beltane being celebrated in winter #wales being portrayed as a completely separated land from england (i wish) #characters getting up after weeks of bedrest like that dosnt completely fuck you up

violetfairydust: #i once read a fic where the flight time from london to seattle was 3 hours