the worst part is steve rogers WOULDN’T. he wouldn’t leave sam with the responsibility of the shield without being there to support him. he wouldn’t go back to a woman who died of old age, had her own life and told him to move on. he wouldn’t have ever, not even once, considered leaving bucky — aka his entire world wrapped up in one person — alone, especially after just getting him back. and he wouldn’t have decided that he’d fought the good fight enough and retire in suburbia in the decade epitomes for traditional values aka an antitheses to everything he stood for. the real steve rogers would legitimately hate the man marvel put on the screen in endgame. and yet. and yet
The conversation around abortion shouldn’t be “are you absolutely sure you want an abortion” it should be “are you absolutely sure you want a child”. You can get pregnant again. You can adopt. But you can’t half-heartedly raise a child or change your mind midway through parenthood. Children are a huge responsibility and if someone isn’t 110% sure they are willing and able to do it, they shouldn’t. Having a baby shouldn’t be the default because of how extremely demanding and difficult parenthood is and the irreversible damage it does to a kid to be raised by someone who didn’t even fully want parenthood and wasn’t prepared for that level of responsibility.
That’s why I take being an aunt so seriously: because I know and value how important it is to take care of a kid, and that I don’t have the resolve in me to do it 24/7.
"just recapping our conversation today, [Company Name]'s policy is to require employees to ask customers about their genitals before allowing them to enter the restrooms" is really not something management wants to see in writing. create a paper trail! be annoying!
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.
In case you just skimmed the post above and missed it, I want to reiterate and highlight Gebru's current position as Executive Director of Distributed AI Research Institute. If you're curious about what AI technology might look like when not applied in the horrifically unethical and damaging way it's currently applied, please check them out.
If we want to have nice things, decentralization is essential, and if we want to decentralize, we need to have our eyes on things that are beyond the scope of the current Big Tech narrative.
The Distributed AI Research Institute is a globally distributed organization of academics, activists, and engineers conducting community-roo
Why are you using chatgpt to get through college. Why are you spending so much time and money on something just to be functionally illiterate and have zero new skills at the end of it all. Literally shooting yourself in the foot. If you want to waste thirty grand you can always just buy a sportscar.
I’m really starting to think you people don’t understand what university is for. You’re buying the accreditation that you can do these things. It doesn’t matter how you do them.
I can assure you if you're going to school to be an xray tech or a surgical assistant it does very much matter how you do the stuff your accreditation says you can do. We aren't all business majors.
Yes, but you actually can’t do an X-ray without an X-ray machine and you can’t do surgery without scalpels. We already rely on technology for everything. Offloading cognitive tasks just frees us up to do more. If you can do your job with chatgpt, but can’t without, you can still do your job. I’m sure you would find university much much harder without access to google or the internet too.
Do you think scalpels are magic and do a little song and dance and perform the surgery themselves like Beauty and the Beast characters and the surgeon is there to conduct the background music
What do you think will happen when your employer, who hired you because they saw you have a certificate to say that you have specific skills and knowledge, starts expecting you to have and use those skills and knowledge and you can't because you think a university degree is just a piece of paper that you buy
"Offloading cognitive tasks just frees us up to do more"
When you're in school, the cognitive tasks are there for the explicit purpose of being brain exercises. It's weightlifting. It is FOR building your mental muscles and making you a stronger thinker and planner. "Offloading the cognitive tasks", then, is just Not Doing The Weightlifting. What happens when you pay for your gym membership and just stand around messing around on your phone? Nothing. Nothing happens. Just money leaving your wallet. Nothing else.
Using AI is a short term pleasure that is going to fuck you over in the long term, and by the time you realize that you didn't build the necessary muscles you need for the cognitive tasks required of your ACTUAL JOB (or, like, adult life in general), it's going to be too late to do anything about it... except going back and doing the real work all over again to get you up to speed.
And if your response as a college student is "Ugh i'm already good at this though, i don't need the practice" -- sweetie, you have no idea how good at it you could be though. If you're good at it now but you keep working on it, you're going to ASTONISH yourself in a couple years with how good at it you can get. I was a good writer when I was in college; I am an ASTRONOMICALLY better writer now, because I put in the work. But you have to lift the weights and build your muscles to get there, even when it's tedious. There aren't any shortcuts for this. You can be content with your own mediocrity, or you can believe that you're capable of growing towards brilliance. Which one will you choose, mediocrity or brilliance? You get to pick right now.
Sometimes homeless people will be off-putting and rude and demand more of your attention than you're comfortable giving. I promise you, if you had to deal with a fraction of what they do on a daily basis, you'd become off-putting too. Sit with your discomfort and direct your anger at the systems which made them homeless, not at the person in front of you who is struggling.
Honestly I miss the energy Tumblr had when the first Pacfic Rim movie came out. Everybody was talking about who they were drift compatible with and that was like a huge compliment. Ppl were drawing Kaijus, and Jaegers, naming ‘em. It was a better time
This doesn’t show exactly what the caption suggests it shows.
In this scene, the lower pilot is dying. He had been captured, managed to escape, and stole a German plane to fly back. The upper pilot–his best friend and rival for the love of Clara Bow*–shot him down, believing he was the enemy. This is him kissing his friend goodbye.
“But that’s still slashy!” you can say. Yep, it is. “You can read this as homoerotic!” Yes, you can. “Why are you denying this? Is it because you think being gay or bi is shameful?” A thousand times no. I am pointing this out because I think this is an important piece of evidence about what homophobia has done to our society and to male expressions of emotion.
In 1927, the obvious reading of this scene, for audiences, was not that this was a romantic kiss. Audiences primarily understood this as an expression of friendship and love, because of course it was perfectly natural for non-romantically involved men to embrace or even kiss, particularly at highly emotional moments. Of course a dying man would want to be held during his last breaths. Of course a guilt- and grief-stricken man would want to kiss his friend goodbye.
However, not very long after this, the commercialization and commodification of homophobia became a powerful force. The market (including Hollywood) began drawing lines and graphs and boxes, declaring which emotions, expressions, habits, and even colors “belonged” to men and to women. This kind of touch, which would not necessarily have been sexualized during many eras or in many cultures, became forbidden to men in the US, Britain and Canada (and many other places, too) within the decade–and is still lost to them today. This scene–a far more honest expression of grief and affection than anything we’re used to seeing in today’s action films–became gay.
Now, if you strongly wish to write “Wings” slash, you can still do so–and not entirely by putting on your goggles! University culture of the 1900s-1920s definitely allowed for a far wider range of sexual behavior than frats do now, etc. I don’t want to police what anybody can and does find in “Wings.” But I think we should acknowledge what we lost when capitalism decided that, for men, kisses could only be sexual.
*You may recognize Clara Bow from that goddamned photo that keeps making the rounds of the internet captioned, “A sex ed class in the 1920s!” so everyone can hoot with derision at the shocked girls in their desks. The photo is actually a still from a movie, and the star, Ms. Bow, is front and center.
“The widely circulated timeline created by @Zerflin does a great job in showing how recently slavery & segregation occurred & that they lasted longer than the modern era.
“I'd like to offer this timeline as another way of viewing the same period of history to show the constancy of both Black resistance in US & efforts of the white power structure to maintain racial caste since 1619.”
I had the exact same experience the second time I looked at this picture as the first time. I was looking like "what is this green line? Like suddenly everything is OK? It's not. Racists are still trying to push us back to 1619. Nothing has been fixed. We still need to fight. Hard!
Life is pain, Highness. @jadeismyname - Tumblr Blog | Tumgag
