Random thing for people to consider is that since Laika is the saint of one way trips should Felicette be known as the saint of safe landings since she did make it back to the ground safely
tu LANCES félicette ? tu lances son corps comme la fusée ? oh ! oh ! prison pour les scientifiques ! prison pour les scientifiques pendant Un Mille Ans !
I'm so glad that that truncated fucking ran-into-a-wall-at-speed tadpole-ass looking squirrel only lives in high altitude forests in Borneo bc this means I am extremely unlikely to encounter one in my day to day life. thank god
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 need to stop replying to “how do you make friends in your 30s?” threads because all my answers boil down to “you have to want to know people instead of have friends” and I don’t think people wanna hear that
It’s like. People can tell if you don’t really like or connect with them. If you aren’t truly enamored with someone you will have a hard time coming up with activities to do together to deepen the friendship. Because you don’t really like that person that much.
i think we should be ridiculing them more for this. you don't get to try and go all "queer website" when your staff likes to go on nuking sprees targeting the trans fem users
Highly unlikely to be of use but I want a card I can carry in my wallet that says “If I Die First In A Traumatic Survival Scenario I Want You To Know I’m Okay With You Cannibalizing Me, It’s Fine, Don’t Be Weird And Guilty About it”. And then on the back there could be like. A list of recipes n shit
To be clear, I don’t WANT to be cannibalized. It’s not my personal preference. It’s just that if I beef it out in the open somewhere and some poor starving fuck has to have a crisis of ethics over using the protein or dying in a hole, I’d want to some way to let them know my dead ass is rooting for them
shoutout to the woman from my high school martial arts class who liked to get me in joint locks and then joke about how I was easy to catch. you cannot comprehend how psychosexually formative that was for me
imagine, if you will, having an adolescent half-crush on someone way older than you, which is also confusingly blurred up with admiration of them as a role model. now imagine that you and that person are in a social environment where it is acceptable to (platonically, consensually) choke someone. I think I was very normal about it considering the circumstances
she would demonstrate takedowns on her husband (also in the class, and who was not a small man) before we got to try them and the first time I saw her twist him around and down onto the floor like it was easy my entire abdomen clenched
I cannot stress enough how eager this guy was to be manhandled (womanhandled?) and flipped around by his wife. he was her de facto guinea pig whenever she got to teach and I never saw him unenthusiastic about it. he'd set himself up for a joint lock fully smiling. the other adults in the class occasionally teased him about it (being so quick to let your wife put you in a submission hold tends to raise a few eyebrows), and I always kind of wanted to defend him but what would I have said? like, don't worry. I won't judge you. I also like being pinned down by your wife
The oldest living tree ever found was a pine named “Prometheus.” It had been alive since before the Egyptian pyramids were built. Some guy cut it down in 1964. Source
he was actually a forestry graduate student who was doing research on bristlecone pines (Pinus langaevea) and got his increment borer stuck in the tree. this tool costs almost $800, so he asked the forest service if he could cut down the tree to recover the tool. after cutting it down, it became apparent that the tree was actually the oldest living organism. ever. (around 8,000 years old). so, not just some asshole. the guy feels extremely guilty and has even broken down in tears during an interview about the accident
So after the grad student cut down the tree and discovered it was the oldest tree in the world he quit studying forestry and went to study salt flats (can’t cut down the oldest trees in the world on salt flats no siree none of that happening) and he was being interviewed about his research, but in the middle of the interview the reporter just stops and says “wait aren’t the guy that…”
And he just takes off running. Literally. Turns around and runs across the salt flats away from the interview and I feel so bad for him but I can’t help but start crying I’m laughing so hard about it
imagine a guy high tailing it across salt flats away from a dude with a recorder
its so different to know it was an accident and that NO ONE was aware until after. its not like this was one ignorant guy cutting down a fucking relic.
"The Montana court separately declared that transgender people constitute a suspect class under the state's equal protection clause. In legal terms, a suspect class is a group that has historically faced such severe discrimination that any law targeting them must meet the highest level of judicial scrutiny to survive—the same standard applied to laws that discriminate on the basis of race. [...] The practical effect is sweeping: any Montana law that singles out transgender people will now face strict scrutiny, meaning the state must prove the law serves a compelling interest and is narrowly tailored to achieve it—a standard that laws almost never survive.
"Because the decision rests entirely on the Montana Constitution, it is insulated from the U.S. Supreme Court. Under the principle of adequate and independent state grounds, the federal Supreme Court cannot review a state court's interpretation of its own constitution, so long as that constitution provides more protection than the federal one. [...] What this means in practice is that Montana's transgender residents now have a constitutional shield completely independent of the Supreme Court of the United State’s decisions."