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.
I’m working on an audio transcript using voice recognition technology, and this gentleman has a very nice accent, but when he says “got” the word is often noted down as “God”.
We don’t know what God tested and what God registered as true or untrue.
We don’t know what God entered into the code since the last time we tested.
We don’t know what God ticketed as an issue and what just God ignored.
Now we know what God changed, but we don’t have a record of what God approved.
HEARTBREAKING: friends who i should be going to the movies and playing dnd and watching anime and cosplaying and going to the mall and having sleepovers and exploring the woods with live one hundred trillion miles away
i don't remember them playing mahjong but they do other old man things like going to the wet market together and drinking soup and taking walks. anyway go watch suk suk / twilight's kiss
Okay, today is Speak Your Own Language Day, so I should be speaking Spanish, HOWEVER I want to use this day of language learning and appreciation to explain something about how Spanish works, and Hispanic people already know that so it wouldn't make much sense to explain it in Spanish, which is why just for this post I'm gonna use English.
I'm gonna be talking, of course, of grammatical gender, because of this viral image:
I'm sure you've seen an image like this floating around and people crying about how having non-binary be translated with a feminine and masculine form depending on usage defeats the purpose of the term.
But it doesn't!!!
Spanish is not like English, it has ✨grammatical gender✨ which has nothing to do with gender identity whatsoever. It's not that we believe chairs are female and stools are male, our grammar is just like this. Every word has a grammatical gender and there must be grammatical gender concordance. Thus, non-binary must have both a feminine and masculine form to use depending with which word you're pairing it.
Let's say you want to talk about a "non-binary person". 'Person' in Spanish is 'persona' a grammatically feminine word (despite its usage being gender neutral and encompassing people of any gender, as I said, grammatical gender ≠ gender identity). Because 'persona' is a grammatically feminine word, you have to apply grammatical concordance accordingly, and so to say "non-binary person" you would say "persona no binaria".
Now let's say you want to talk about "non-binary gender". 'Gender' in Spanish is 'género' a grammatically masculine word, and because of that to say "non-binary gender" you would say "género no binario".
See? It's not about grammatically imposed misgendering, it's about how this language is built.
But Shine, I hear you say, that's all nice and good, but how do I refer to my non-binary friend? And well, dear reader, you're in luck because I'm not only Spaniard but also non-binary myself.
In Spanish 'friend' (like a whole lot of words) has a feminine and masculine form. When those words are used to refer to people, grammatical gender does match gender identity. For the most part. And broadly speaking, feminine words are associated with the vowel -a, and masculine with the vowel -o (this is not universal, there are exceptions to this, I'm trying to paint the broad picture to give you a general idea).
So what about gender neutral? Well, officially we don't have one. The Real Academia Española doesn't recognize it... But the RAE isn't word of god, it compiles usage, so the more a term is used, the more chances it will be officially recognized. Not using a term because it's not officially recognized is actively detrimental to the goal you want to achieve.
Okay, not an official gender neutral, but what have we come up with? Well, at least in Spain, it's associating gender neutral to the vowel -e. So you have the femenine ella/la/-a, the masculine él/lo/-o, and the gender neutral elle/le/-e.
Now time for practical examples!
"My friend, David, is a boy." 🇬🇧 → 🇪🇸 "Mi amigo, David, es un chico."
"My friend, Liz, is a girl." 🇬🇧 → 🇪🇸 "Mi amiga, Liz, es una chica."
"My friend, Alex, is non-binary." 🇬🇧 → 🇪🇸 "Mi amigue, Alex, es no binarie."
And that's how you do it!! At least in queer friendly spaces in Spain, can't speak for other places. I have seen the 'x' thrown around to make gender neutral in Spanish, but -e is way more intuitive for spoken language, so I like it better.
But anyway, thanks for coming to my TED talk and remember I generalized a lot when doing this. If you're learning Spanish I can only wish you good luck in dealing with our bullshit if you come from English, and assure you that once you internalize our orthography rules you will never again mispronounce a word you read for months before you hear it spoken (we have a very consistent spelling/pronunciation system, gotta be one of my favourite things about my language).
jacobvanloon
stormborns
snowghoul
theartofmadeline
gummygunk
blakekathryn
titsay
