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.
This is the second time I've seen a video of this technique and this explanation is so clear! It does use more fabric than English paper piecing (EPP) but you end up with a double sided hexagon so don't have to source fabric for the backing.
I'm doing EPP at the moment but I have a hole punch to make the papers and just use leaflets and junk mail, so it doesn't feel wasteful. I don't think it's difficult either- in the video she mentions it's not for beginners, but I don't have that much experience with hand sewing or EPP and I've been finding it pretty easy so YMMV
I saw this video yesterday and was seized with the need to try it out immediately. Lookit my cute lil' hexagon baby!!
Here is what the backside looks like. OP notes this takes more fabric than paper piecing, but that excess fabric makes it already triple-layered. Besides not needing backing fabric, I don't think you'd need batting for this quilt at all. It's already thick and soft just from folding all that fabric into a hexagon.
Hexagon quilt tutorial video by tiktok user camelscrafts. Method:
Each hexagon begins as a 6" circle. camelscrafts does this by creating a paper template using a compass. According to the video, a 6" circle will create a hexagon that is 2.5 inches tall.
These hexagons are hand-sewn. Thread the needle.
With the fabric right side facing, find the center of the circle by folding it in half right sides together, then folding it in half again (wrong sides are facing). The top of the triangle shape is the center of the fabric circle.
Make a small stitch into the center of the fabric. The wrong side is still facing.
Unfold the circle. There will be a small stitch in the center.
Now the hexagon is created by folding the circle into itself: Take the needle to one of the edges of the fabric (it doesn't matter which one). Pull the needle through and pull the thread tight. This will fold down the fabric and create an edge of the hexagon. Crease the fold with your finger.
This fold has two corners, one at the top and one at the bottom. Put the needle into one of the corners and pull the thread taut. This will create another fold.
Continue this going around the circle until all of it is folded down, creating the hexagon. camelscrafts notes that the last corner pulled in may be a little bit "wonky" (no precise point in the corner) if the corners were not done precisely. However, that corner is pulled into the back, so is not visible from the front.
The hexagon is now formed. Sew around the folds in the middle of the circle to hold the folds in place. Tie off and cut the thread.
Attach hexagons to each other along the sides. With right sides together, whip stitch the sides together.
that's a very good addition actually, a solid "you want me to ask people's kids about their genitals? can i have that in writing?" should make upper echelons very uncomfortable.
I always thought she'd be sent to the Agricultural Corps, but now I see why she wasn't. It's not that Master Yoda feels sorry for her. It' that she's already passed the test the rest of us will be facing — (Whie, about Scout.)
✨🌙 ART LOG -> @404ama
The comic text is from '3gatsu no Lion' it has been in my Jedi-isms folder for years I was so excited to use it with this theme T_T
From the Nashville Zoo’s fb page! Here’s the petition, please please please take a moment to add your name (even if you’re not from Nashville!). If you are from Tennessee, contact your representatives and make it clear that the people do not want this data center. This is an AZA accredited zoo which is home to several species of critically endangered animals, we NEED to protect it. Make your voice heard!
90-ghost has done so much for the people of Palestine. He is trying to get the rest of his family out of there while the war rages on, so to show your support, please donate to his sister's and brother's respective Paypals!
Help support Samar Saed Kamel Saed by donating or sharing with your friends.
Go to paypal.me/bushrabo and type in the amount. Since it’s PayPal, it's easy and secure. Don’t have a PayPal account? No worries.
@gazavetters Vetted List lists Ahmed and his family at #740.
Happy (belated) 32nd birthday, Ahmed. I hope you know how much you've done to help the Palestinian people. I might not have any money, but I shall always do my best to spread your hard work and the words of others.
Police take fiscal oxygen out of the room, away from actual social support programs. The money spent on policing could go to housing, nutritional services, healthcare, civil rights, building inspections, labor rights enforcement, street maintenance, public transportation. But it goes to paying guys to sit around idling Ford Explorers all day and hitting poor people with sticks.
When Amal surprised me by telling me she named her newborn after me, I felt like part of my heart had gone to Gaza and remained there ever since.
This part of my heart is now breaking. Amal, who's been on a poor diet because she can't afford better food, is unable to breastfeed baby Mina any longer, and is unable to afford enough formula to compensate, and now both mother and baby are weak and malnourished.
Amal is suffocating from the stress of having to provide for her daughters. Her husband Motasem is doing everything he can (recently, he was almost caught in a bombing while trying to get food) but they're both overwhelmed. I wish there was anything I can do for them.
I'm begging you to donate if you can and share if you can't. I want, more than anything in this world, for these kids to grow up healthy and for their parents to be able to rest.
Credit card companies will TRY to saddle you with this kind of debt by the way - if ever a loved one dies and you are not co-signed on their credit card, do NOT agree to pay their debt unless you ask a lawyer first if you truly have to.
They will say “don’t you want them to go to the grave without debt”, they will try to guilt you, they will take advantage of your vulnerability.
Source: when my father died, he had some credit cards that my mom wasn’t on that she had no access to. The companies contacted her while she was sorting through the bills and getting a handle on how to run the house alone, badgering her with his credit card debt.
She wasn’t liable for any of it, but if she had ever agreed to pay before finding out that she didn’t need to, she would have been considered to have taken on his debt and would have HAD to pay it. It’s slimy, it’s predatory, and it’s entirely legal for them to do this.
Never accept the credit card company’s word about your obligation to pay anyone else’s debt, if you don’t have access to the card, ask a lawyer before agreeing to anything.
theartofmadeline
kaledo-art
discoholicmusic
pixelskylines
janainaart
jacobvanloon
2087
