Im enjoying the longevity of tumblrs recontextualization style of humor. a seemingly innocuous post followed by like "posts that a gnome would make" or like "are you a phone"
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.
So my beta reader for the Big Fics is an astrophysicist, right. Who is currently also writing a hard sci-fi novel about the exploration of Phobos (more power to them, I cannot with the physics required for that, best I can do is soft sci-fi/fantasy and that reminds me I should finish that story).
Anyway I was bitching about how hard it is to come up with feasible planets in Star Wars because sometimes you need a new planet from scratch and sometimes you need to know more about a planet than the 'has jungles, is probably a moon technically' than Wookieepedia will give you, and they're like 'oh yeah I can do something about that'.
So they've written (in Matlab but they swear it will run as a .exe as well and I may be conscripted to embed it as a web tool at some point) a star system generator.
You input what you know about the planet (ecosystem, population, sun colour, does it have liquid water, does it have a moon or moons, is it a moon or moons, temperature averages, atmosphere, you get me) and it will give you the... everything else about the star system, in obedience to real-universe physics. And if you input nothing you get a randomly generated star system.
And I’m like oh I know people who will be into this with a vengeance, and they're not on Tumblr, so this is me seeing who exactly would be keen on, and I cannot stress this enough, a real-physics comprehensive star system generator.
It's still in the debugging phase (last error fixed: every planet wants to have a population of exactly 5000 regardless of other factors, turned out to be a missing equals sign somewhere), but I'm psyched for this and trying to gauge interest for how high a priority 'make this an accessible web tool' needs to be.
I love how varied and universally weird the circumstances for making lifelong friendships are. Here's this guy I accidentally messaged once and I could not imagine my life without them now. Here's this girl I was so scared of when I met her, I would kill for her and remind her to rest on the regular. Here's this other guy we have so much in common we used to joke we were the same person in different timelines. It took us years to meet in person and I attended his wedding. There are also people who entered my life in absolutely unremarkable ways but changed it forever for the better. It's wonderful how easy it is to find people to love.
rbing again real quick--i and other(s?) have noted in other rbs that this is almost never done. it's for very few properties, and it's only for timestamped warnings that are done in-house, not by users.
there are still timestamps available for warnings on those same triggers for those same properties. just bc there's a paywall on something doesn't mean other comments and timestamps can't be posted or read for free on that property or that trigger.
see here:
i don't have an account and haven't paid for anything on the site, so all of those comments are what i'm seeing for free.
i had to search and search to find anything that had paywalls, and had to resort to looking at The Substance, as referenced here. i've been using the site regularly since the op posted and have never organically encountered a paywall on anything, including on very popular properties.
i don't see any reason to stop using the site--everything is still available for free. they just have added another way for some ppl to choose to pay some $ (in addition to ads), since this is a completely free service, and no website is free to maintain.
as an example, someone with the money could pay to see the timestamps, and then leave a free comment with the same timestamps for others if they wanted to.
I've seen this post on my dash a few times now and I want to boost this specific addition. I've emailed the owner of the site to get clarification on this, and I was told the motivation behind this. He wanted to start paying some people to memorize all 200+ triggers on the site, sit down and watch popular movies/shows to thoroughly analyze whenever any trigger pops up, and note down timestamps and details of them. Ads unfortunately do not cover this (they probably only cover the site's base upkeep), so the subscription service is specifically to pay these people. Community comments have never been paywalled and are still accessible.
He thanked me for my feedback when I expressed confusion and concern about the paywalled comments being mixed in with the free comments, and has since updated the site to make them clearly distinct, along with a preview of what the paid comments look like:
He was very transparent about it all and very kind. He also said this was his mistake and is trying to clear it up. Please continue to use DoesTheDogDie.com, it is an excellent resource and there is nothing else like it!
"There are legends of people born with the gift of making music so true, it can pierce the veil between life and death. Conjuring spirits from the past and the future."
SINNERS (2025) dir. Ryan Coogler
kianamaiart
germansingergirl
baratheonpilled
izzy806
andulkaphoto
blakekathryn
vaporwave
