before my egg cracked, i had noticed that trans people were often pro-accessibility and up-to-date on the needs of disabled people, but i hadn’t seen any inherent connection between the two (other than the obvious minority-looking-out-for-other-minority thing). but now that i’m trans and medically transitioning, and i have to constantly repeat myself while talking to doctors and nurses, and explain things about my own anatomy to medical staff who should already know this, and having every single problem i might have blamed on my “condition” so nothing i say is taken seriously, all of the sudden i have a little sneak peak into the life of someone who has to deal with this all the time. like shit bro, being disabled probably sucks ass, someone should do something about this
Something by I love about Black Menswear is how they’re not afraid of color and personality. It’s not just the same boring black or blue suit jacket every time. It’s also just very dapper.
they've been arguing about malgendering on twitter so here's the truth from me, the arbiter of truth (jk but ive been talking about malgendering long before it became a popular buzzword and heres my understanding of it)
malgendering is a form of transphobia wherein a transphobic person uses their victim's correct gender to harass or demean them. it is meant to be the inverse of misgendering, where a transphobe uses the victim's incorrect gender to harass or demean them.
misgendering towards trans women can be as simple as "youre a man." but malgendering is more insidious, like someone saying "you really are a girl" when she makes a mistake in order to insult her intelligence. the phrase "you really are a girl" is positive when removed from the context of the trans woman making a mistake and the transphobe's sarcastic tone that implies shes stupid in an misogynistic way.
if a trans man does something like standing up for himself against an unfair situation or asks for better accommodations/treatment, and a transphobe wants to use that to imply he is selfish and entitled, they might say "trans men really are men" which again, is a true statement on its own. but in the context of the situation, its clear the transphobe is using the stereotypical worst qualities of men to insult and undermine someone's very identity.
malgendering can be any severity of harm, just like misgendering. my mom slipping up in the first months of my transition and accidentally calling me "she", and a bigot shouting "YOURE A WOMAN YOULL NEVER BE A MAN" at me are both examples of misgendering, but one is much worse than the other and its the same with malgendering. being told "trans men are the men of the trans community" as an insult is annoying and rude but not a big deal. however, right now ICE is torturing trans men in concentration camps by forcing them to do pointless and grueling labor. and when they become exhausted, theyre told "I thought you were a man. if youre really a man this should be easy for you" these are both examples of malgendering. one is a microaggression, the other is literal torture.
malgendering happens equally to both trans men and trans women, its not something that is unique to anyone because it is a form of transphobia. the strategy is that if a transphobe tries to make a trans persons life as their true gender so miserable and unsafe, detransitioning seems like the safer and happier option. its psychological torment, moreso than the plain and obvious insult of intentional misgendering. so dont let it happen, call it out when you see it and dont be fooled by transphobes trying to erase the context and be like "what!! what i said was true!"
For my own understanding, let me try to boil down your explanation into a one sentence definition:
Malgendering is the action of maliciously burdening a trans person with the stereotypes about their true gender, with the intent of making living as said gender seem less favorable.
Does that catch what you're trying to say? Because from your post I'm not sure of the intent bit.
yes! the transphobic person may not be aware of their intent and not even know what they're doing, most of the time theyll deny that they are being tranphobic because they think theyre 'affirming" someones gender. it serves as an easy way out and often malgendering is more like a dogwhistle, only some people can hear it. but yes, that is the "purpose" of malgendering.
and you could argue "well thats just sexism. treating a trans woman like shes stupid or slutty because shes a woman is just misogyny. treating a trans man like he must be entitled or predatory because hes a man is just misandry" yes and no, because the context matters.
saying these things to a trans person with the knowledge that theyre trans has an underlying implication: "you wouldn't be bad if you weren't trans". when you tell a trans men that all men should die and you hate all men, youre telling him "you wouldnt be evil if you detransitioned. everyone would be kind to you if you were still a girl" and same with trans women: "you would be smarter and respected if you weren't a woman. I wouldnt harass you if you were still a man."
malgendering says "these are the consequences of being a man/woman, and im going to make sure you suffer them."
I am sort of frustrated about that post I made about Google Gemini giving an AI summary of Tumblr users if you google a URL because it quickly turned into a long post where people started sharing all the things the AI got wrong about them, and I'm not saying it isn't good for a laugh when an AI is wrong, but I feel like people are missing the point. Who cares if the AI offers information about you that's wrong? What's worrying is all the information the AI got right.
That initial accidental google I did of a mutual returned an AI summary that was absolutely correct, and featured information about them that would not have been obvious from a quick perusal of their blog. The first few times I googled myself the AI overview was 100% correct. The version of that post that is circulating is not the one where I shared that the AI gave me someone's real name. I had hoped that it had screwed up and it wasn't their real name after all, but no, I checked with them, and just by googling someone's Tumblr URL, I got their honest to goodness legal name. Do you have any idea how dangerous that is? Like yeah I get that it's fun to go, "Haha, AI bad at collecting information," but that's not the problem. The problem is that sometimes AI good at collecting information, and it can collect a lot more a lot faster than a normal human can, by running multiple searches simultaneously and cross-referencing a lot of different documents all in a few seconds, for a level of snooping that would take a real person hours or even days, if they could do it at all.
That is so dangerous and I need people to be angrier about the fact that Google thinks it's okay to let their AI do this for online users. If you google your legal name and you're not already famous the AI won't say shit. Because that would be creepy and invasive! So why does Google think it's all hunky dory to do it for random internet users?
If you're reading this and thinking that sounds incredibly scary/unsafe, please go into your blog settings and make sure the setting to discourse external searching is turned on!!!!
^ it should look like this ^
I've had that setting on for a long time. When I saw this post I tried googling my account and no AI summary appeared. I decided to experiment, and I turned that setting off and googled my url again–that time the AI summary appeared. It got a lot wrong, but it did get some things right, and had a concerning amount of information considering a few minutes before the AI summary didn't even appear. It also had the state I live in listed, accurately. That's not something I'm super secretive about, I have said it on here before, but it's also not something I necessarily want an AI telling someone umpromted just because they Googled my URL–frankly I think it's absolutely insane that the AI is able to do that without people questioning it. I quickly turned the discourage external searching setting back on, and I'm no longer getting the AI overviews. But it's definitely something we all should be aware of–especially on other social media sites that don't have that feature.
[ID: screenshot of a setting turned to the on position. It reads:
Discourage external searching of apollomission8
External search engines like Google or Yahoo will be discouraged from indexing your blog and showing it in search results. It is ultimately up to them to honor your request.
they've been arguing about malgendering on twitter so here's the truth from me, the arbiter of truth (jk but ive been talking about malgendering long before it became a popular buzzword and heres my understanding of it)
malgendering is a form of transphobia wherein a transphobic person uses their victim's correct gender to harass or demean them. it is meant to be the inverse of misgendering, where a transphobe uses the victim's incorrect gender to harass or demean them.
misgendering towards trans women can be as simple as "youre a man." but malgendering is more insidious, like someone saying "you really are a girl" when she makes a mistake in order to insult her intelligence. the phrase "you really are a girl" is positive when removed from the context of the trans woman making a mistake and the transphobe's sarcastic tone that implies shes stupid in an misogynistic way.
if a trans man does something like standing up for himself against an unfair situation or asks for better accommodations/treatment, and a transphobe wants to use that to imply he is selfish and entitled, they might say "trans men really are men" which again, is a true statement on its own. but in the context of the situation, its clear the transphobe is using the stereotypical worst qualities of men to insult and undermine someone's very identity.
malgendering can be any severity of harm, just like misgendering. my mom slipping up in the first months of my transition and accidentally calling me "she", and a bigot shouting "YOURE A WOMAN YOULL NEVER BE A MAN" at me are both examples of misgendering, but one is much worse than the other and its the same with malgendering. being told "trans men are the men of the trans community" as an insult is annoying and rude but not a big deal. however, right now ICE is torturing trans men in concentration camps by forcing them to do pointless and grueling labor. and when they become exhausted, theyre told "I thought you were a man. if youre really a man this should be easy for you" these are both examples of malgendering. one is a microaggression, the other is literal torture.
malgendering happens equally to both trans men and trans women, its not something that is unique to anyone because it is a form of transphobia. the strategy is that if a transphobe tries to make a trans persons life as their true gender so miserable and unsafe, detransitioning seems like the safer and happier option. its psychological torment, moreso than the plain and obvious insult of intentional misgendering. so dont let it happen, call it out when you see it and dont be fooled by transphobes trying to erase the context and be like "what!! what i said was true!"
For my own understanding, let me try to boil down your explanation into a one sentence definition:
Malgendering is the action of maliciously burdening a trans person with the stereotypes about their true gender, with the intent of making living as said gender seem less favorable.
Does that catch what you're trying to say? Because from your post I'm not sure of the intent bit.
yes! the transphobic person may not be aware of their intent and not even know what they're doing, most of the time theyll deny that they are being tranphobic because they think theyre 'affirming" someones gender. it serves as an easy way out and often malgendering is more like a dogwhistle, only some people can hear it. but yes, that is the "purpose" of malgendering.
and you could argue "well thats just sexism. treating a trans woman like shes stupid or slutty because shes a woman is just misogyny. treating a trans man like he must be entitled or predatory because hes a man is just misandry" yes and no, because the context matters.
saying these things to a trans person with the knowledge that theyre trans has an underlying implication: "you wouldn't be bad if you weren't trans". when you tell a trans men that all men should die and you hate all men, youre telling him "you wouldnt be evil if you detransitioned. everyone would be kind to you if you were still a girl" and same with trans women: "you would be smarter and respected if you weren't a woman. I wouldnt harass you if you were still a man."
malgendering says "these are the consequences of being a man/woman, and im going to make sure you suffer them."
alright I finally got around to watching netflix's "The Dinosaurs"
(Don't become a paleontologist kids, it turns watching dinosaur related things from "fun" into "work")
THINGS DONE WELL
Birds are treated as full dinosaurs throughout
Emphasizing that dinosaurs had complex behaviors and social structures
Plenty of fluffy dinosaurs present
Most of the reconstructions are fantastic
KPg presented as tragedy and not inevitable
The ending showing all the different dinosaur behavior juxtaposed with modern dinosaurs doing that behavior is glorious, ESPECIALLY because it wasn't just theropods
THINGS DONE... NOT AS I WOULD DO THEM
the only named titanosaur is the smallest one??? what are we doing
For that matter, please just name the dinosaurs. I promise, audiences aren't allergic to names. I promise.
stop acting like the Jurassic was the time of giant dinosaurs when EVERY "BIGGEST" DINOSAUR WE KNOW IS FROM THE CRETACEOUS
I'd make more of the mid-sized theropods fluffy or at least have some decorative floof
I'd have more of the stepwise process of the evolution of birds as we go straight from early Jurassic dinos to Anchiornis, including showing how complex feathers evolved for display and communication before flight
can we PLEASE stop using the Kpg for TraumaPorn, I'm not kidding, I'm sick of watching dinosaurs die horrible horrible agonizing deaths for twenty minutes at a time
WE KNOW BIRDS ARE DINOSAURS. they were PRESENTED as dinosaurs THROUGHOUT. the "Gotcha! they didn't all go extinct!" thing at the end doesn't MAKE ANY SENSE when we ALREADY KNOW
also acting like its a Tuatara situation - implying just a small remnant is holding out - when birds cover the entire planet and outnumber mammals a significant amount is... a Choice. (It's a bad choice)
STOP ENDING DINOSAUR DOCUMENTARIES AT THE KPG WHEN WE KNOW THAT ISN"T THE END OF THE STORY
THINGS THAT ARE JUST PLAIN WRONG WTF
what the actual eff was the Spinosaurus sequence
MOST carnivores engage in opportunistic cannibalism. that doesn't make them "a cannibal" and it doesn't make them special, either, because most do it.
I did see quite a few somewhat pronated hands
Dinosaurs were either straight up endothermic or gigantothermic. they could handle snow. they DID handle snow. why do we keep acting like they died the minute cold water touches their skin. we don't have a lot of evidence of titanosaur migration because cold. what is this.
Why is Alamosaurus with Edmontosaurus. If they aren't Alamosaurus, why are we just making up titanosaurs for fun. Like I know they almost overlap but the emphasis is decidedly on ALMOST
hateg island in the non-latest cretaceous segment because...????
anchiornis could probably fly. the gliding stuff is almost certainly dead in the water.
In short, I liked it a lot, but I did yell at the screen frequently too, so what can you do
I just want a "story of the dinosaurs" docuseries that COVERS THE CENOZOIC. I think we've gotten enough out of the "violent end" narrative and we have plenty of documentaries that just cover the mesozoic. make the madness stop. no, we don't have to ignore the extinction, but that doesn't mean we can't do "There was a tragedy... nonetheless dinosaurs rose from the ashes..."
like yes, it's a surprise to HUMANS that dinosaurs are still around. you know who isn't surprised?
THE TWENTY THOUSAND (probable) SPECIES OF DINOSAURS THAT COVER EVERY CORNER OF THE PLANET
THEY KNOW. ITS NOT A SHOCK. WE WERE JUST STUPID.
anthropocentrism will actually be the death of me
Anyways, check it out. Just remember Spinosaurus probably wasn't... That.
1- @Bluewmist on Twitter / 2- Roly Poly is Taken on Twitter / 3- About Time (2012) by Richard Curtis, image from Mita Park on Unsplash / 4- Sherri Turner on Twitter / 5- Cold Solace by Anna Belle Kaufman / 6- The Anthropocene Reviewed by John Green
Flint is so funny because he's an example of anti-system leader who want the best for his fellow man.
But doesn't LIKE his fellow man. Or woman. He tries, but he really just can't help but feel annoyed by most of the people he is trying to lead against the system (or liberate from it) and feels slightly superior to most of them.
He settle on wanting to free his fellow man or woman from an evil Empire because he likes and respects England even less than he likes and respects his fellow man.
Which is valid btw but it's fun how he genuinely only likes and respects a max of five people in the whole show, and yet he still says things like "I can't believe we are so poorly made as that."
Yes you believe that in your daily life, that's why you can't open up with anyone and you can’t stop treating them as your politically, morally and intellectually inferiors!
I think it's fascinating that in abstract he believes in the inherent goodness of humanity and the idea that in a more equal and less corrupted society, people would not oppress each others for fun and history would be a journey though progress; in concrete he thinks that every person he interact with is dumber than him and need to be manipulated and convinced to do exactly what he thinks is the right thing to do, with very fews exception of two or three people he considers as smart and morally righteous as himself.
What a funny character.
The contraddiction of his character between his political/ideological optimism about humanity as a whole and his deep distrust and dislike of almost any individual he meets on a personal and intellectual level is so entrateining and very realistic.
These pescatarian birds are directly exposed to PFAS contamination due to the island's position near the St. Lawrence Seaway.
Over fifty years of data show a peak in PFAS (also known as "forever chemicals") content in seabird eggs in the 90s, followed by a decrease as regulations went into effect. The most recent findings show a 70% decrease of most common PFAS.
While continued vigilance a regulation is needed, this data indicates that regulations are working to reduce PFAS concentrations in marine ecosystems.
Yes!!!! I did a review of literature on PFASs in human drinking water about half a year ago, and there is a lot of really good progress! Please celebrate this, please don't let this solution be forgotten (at least so quickly) as the ozone layer or acid rain.
We are making genuine progress! Producers are dramatically altering how much they use PFAS and how much gets released in effluent, but also there's a lot better understanding of how to remove PFAS from the environment!
@the-no-dont-do-its very good question! firstly, it's important to point out that on their own, they don't. we have to actively apply methods to remove them from the environment. these methods are LARGELY based on adsorption, which is sort of like filtering except it involves the chemical getting stuck to something else (the adsorbing material).
you can think of this sort of like how water wicks into a paper towel. the water gets stuck to the paper because it's attracted to it via capillary forces, even though there's no chemical reaction going on.
the two main methods used are granular activated carbon (GAC) adsorption and ion exchange (IX).
activated carbon is already pretty familiar to a lot of us; it's the stuff in a lot of replaceable water filters. the activated carbon has a huge internal surface area, and that allows for the fairly weak intermolecular forces to add up and allow contaminants to get "stuck" onto the surface of the activated carbon. over time, the activated carbon gets filled with junk, and you have to replace it.
GAC is essentially this, except that the activated carbon is granularized and produced in specific ways to maximize how much it attracts certain chemicals. this can be tuned because activated carbon gets its massive surface area from internal "pores", and various processes will change how large and frequent those pores are.
It's essentially a Russian nesting doll of pores, and controlling the size of the larger pores influences the permeability of the activated carbon and controlling the size of the smaller pores (micropores) influences what exactly is most attracted to the activated carbon.
However, GAC has a few major downsides:
It is not specific to PFAS. This is more of a mixed blessing because it was already frequently used and well understood, and the infrastructure for producing and distributing it already existed. However,
It loses effectiveness over time and must be replaced. This is a continued cost, albeit a low one, but this has one final major issue
As time goes on, the PFAS previously adsorbed to the activated carbon is desorbed and replaced by other things that have a higher affinity for the activated carbon.
As such, ion exchange (IX) was always very compelling. The whole point of it relies on the fact that PFAS molecules are predominantly made of two parts: An acid head group (either a carboxylic or sulfonic acid group) and a perfluorinated tail.
The head groups on the right are what become ionized—or specifically, deprotonated. A hydrogen leaves and is replaced with a metal cation (usually sodium), forming a PFAS salt (chemical meaning of salt!). These are much more soluble in water because of polarity reasons, and so the mobile PFAS molecules are almost always in that salt form.
By passing through these PFAS salts through a permeable polymer matrix that has (1) numerous positively charged groups like quaternary amines and (2) highly mobile negative ions loosely attached to those stationary positive groups (most often chlorides), you can actually get the PFAS to be "stuck" inside the polymer matrix and what comes out is just good ol' sodium chloride, or salt (culinary meaning of salt!).
This shows a version with hydroxide (OH-) ions as the mobile anion, but it's the same idea. The +NR3 in yellow are stuck to the polymer matrix, but the OH- can freely move around. However, without another anion to replace the OH-, the ionic attraction prevents the hydroxides from leaving.
In comes the PFAS. Despite being slightly soluble in water, the anionic PFAS aren't really that mobile, and when they pass through, it's much easier for the hydroxide ions to leave. Another very important effect is that the long perfluorinated tail of the PFAS is attracted to the polymer matrix, whilst the counterions are ONLY attracted via the ionic force. Thus, PFAS would much rather hang out in the polymer matrix.
Of course, IX has its own downsides
These resins are much more expensive, both to manufacture and to transport.
While they can be "regenerated", it's a tricky process that currently requires the use of nearly anhydrous methanol, which is both poisonous and extremely flammable, increasing the operating costs.
As the hydrophobic tail is a key part of allowing the PFAS to stick to the matrix, short-chain PFAS are very poorly dealt with by this system. This is exacerbated by competition between different PFAS molecules, as long-chain ones will cause short-chain ones to desorb.
Overall, the best method appears to be using a series of ion exchange resins followed by an activated carbon filter. The ion exchange will capture the bulk of the PFAS molecules, and the activated carbon will grab any stragglers. Effective filtering of other contaminants prior to the PFAS removal system will also ensure minimal competition in the activated carbon.
And a SIGNIFICANT amount of this understanding has come in the last fifteen years. In particular, the idea of ion exchange is very new! Twenty years ago, it was seen as WAY too expensive, fragile, and ineffective to ever be a useful technology. Nowadays, it's widely implemented in problem areas and we've built up the infrastructure to support it.
Reblogging this because of the really excellent explanation!
Also a welcome reminder that just because a Big Problem doesn't seem practically solvable right now doesn't mean it won't become solvable in the future!
Twenty years ago ion exchange wasn't a feasible solution to PFAS because of cost and logistics, now there is infrastructure to support it. Twenty years ago solar panels were seen by many as too impractical and expensive for large scale energy generation, now they're the world's cheapest source of energy.
The Burger still gets made, even if you go Vegan. If you don’t buy it, it just winds up in the trash. If you want to do something meaningful about waste, you need legislation: It must become a crime to waste food in those ways.
If you care about Animal Cruelty in Factory Farms, you need to get legislation passed. It must become a crime to mistreat animals in those ways, and when malfeasances occurs, the onus of responsibility for those crimes must fall upon wealthy shoulders. That, also, requires legislation. It requires regulations, and regulators.
The largest source of Microplastics is wear and tear on automobile tires. It doesn’t matter what brand of shampoo you buy. It doesn’t matter which company you support with your dollar. The issue of Public Transit is too large-scale to be handled at anything less than the municipal level.
It’s not enough to just not participate in society
If you want the world to Change, you must leverage the mechanisms of political power.
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.
