we justifiably give Biden a lot of shit but I think "at least 3" is the funniest possible response to some right wing dipshit asking you how many genders there are
Taking this a little too seriously, the real answer kinda depends on what is meant by "gender". But funnily enough from a purely biological perspective, Biden's answer is pretty much on exactly point.
Even if we categorize biological features only as either purely male or purely female, sometimes it can happen that someone is born having both some male and some female features. That's no longer purely either, necessitating a third thing. This is called intersex.
However depending on what you wanna count as gendered features, different intersex people can have different combination of those features so an argument can be made that each combination of those features can be considered its own gender.
So yep, biden is not just being funny here he pretty much gave the most correct answer he could have: There's 3 genders minimum, possibly more depending on your definition of gender, and this is neither the place nor the time to argue about what exact definition we're talking about.
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.
People should say "thank you for asking" more often.
To encourage people asking nicely instead of doing something without asking or not doing or learning anything at all and be sad about it.
Too many "yes" in someone's lives can make someone feel like they can just do it anyways, to many "no" can make them feel why even bother asking or wanting something at all.
But the phrase "thank you for asking" always reminds someone that the act of asking is often just as important as the answer.
Can't have people getting along if they keep getting in each other's way unknowingly because no one ever asks the other.
Not that anybody asked, but I think it's important to understand how shame and guilt actually work before you try to use it for good.
It's a necessary emotion. There are reasons we have it. It makes everything so. much. worse. when you use it wrong.
Shame and guilt are DE-motivators. They are meant to stop behavior, not promote it. You cannot, ever, in any meaningful way, guilt someone into doing good. You can only shame them into not doing bad.
Let's say you're a parent and your kid is having issues.
Swearing in class? Shame could work. You want them to stop it. Keep it in proportion*, and it might help. *(KEEP IT IN PROPORTION!!!)
Not doing their homework? NO! STOP! NO NOT DO THAT! EVER! EVER! EVER! You want them to start to do their homework. Shaming them will have to opposite effect! You have demotivated them! They will double down on NOT doing it. Not because they are being oppositional, but because that's what shame does!
You can't guilt people into building better habits, being more successful, or getting more involved. That requires encouragement. You need to motivate for that stuff!
If you want it in a simple phrase:
You can shame someone out of being a bad person, but you can't shame them into being a good person.
I came up with this three-way table to help me (and now you, if you want) to rate things out of 5 stars. I was thinking of books and films when I made it, but you can probably use it for other stuff.
The idea is that you rate the thing on how much stuff you loved and how much stuff you hated, and those things weight against each other. There's only one way to get 5 stars or 1 star, so those should end up as the rarest ratings, wtih 3 stars being the most common.
'Spicy' means that the thing inspires emotion, whether positive or negative, while 'bland' means it doesn't affect you much either way.
An example of a 3-star (spicy) - for me personally - would be the Twilight series, because there's plenty of garbage in there but also some things that are like crack to me. I can't think of an example of a 3 star (bland) because by nature they don't stick in the mind.
(This also assumes giving 0 stars isn't allowed. That'd throw it out of whack...)
being sick & miserable objectively sucks, but it has become significantly easier to cope with since learning that “sickness behavior” is a well documented part of the body’s immune response
feeling not only physically but also emotionally like fucking garbage is unfortunately an extremely effective way to force your body to prioritize fighting infection & keeping you alive. i don’t have to like it, but knowing why i get weepy & pathetic when sick does help at least a little
It's worth noting that the reason why nearly nearly humans have a universal response to sickness is because across the entire history of our species the ones that didn't kept dying.
Also it's not just to help our own bodies to recover from disease but also to limit spreading the disease to others. Most if not all of the parts of sickness behavior also result in reduced contact with other people.
Limiting the spread of disease is actually the most important part of paid sick leave. You can't have a functioning business when half of your workforce is dying because the sick guy just absolutely needed to be at work.
These dudes are fucking legit. They don’t just show up one day in court, either, they actually make friends with the kids and let them know they have a support system and that there are people in the world who care about them and will always have their back. And less important, but also cool, is that the few times a couple of them have come into my cafe, they’ve been super friendly and polite and when I told one of the guys that I noticed his Bikers Against Child Abuse patch and wanted him to know how awesome I thought he was because of it, he got kind of shy and blushed and said, “The kids are the awesome ones, we just let them know they’re allowed to be brave.”
The source is long, but so, so good. These men and women are available in 36 states, 24 hours a day to stand guard at home, in court, at school, even if the child has a nightmare. Many of them are survivors of childhood abuse as well, and know what it’s like to feel scared and alone.
In court that day, the judge asked the boy, “Are you afraid?” No, the boy said.
Pipes says the judge seemed surprised, and asked, “Why not?”
The boy glanced at Pipes and the other bikers sitting in the front row, two more standing on each side of the courtroom door, and told the judge, “Because my friends are scarier than he is.”
Bikers Against Child Abuse was founded in 1995 by a Native American child psychologist whose ride name is Chief, when he came across a young boy who had been subjected to extreme abuse and was too afraid to leave his house. He called the boy to reach out to him, but the only thing that seemed to interest the child was Chief’s bike. Soon, some 20 bikers went to the boy’s neighborhood and were able to draw him out of his house for the first time in weeks.
Chief’s thesis was that a child who has been abused by an adult can benefit psychologically from the presence of even more intimidating adults that they know are on their side. “When we tell a child they don’t have to be afraid, they believe us,” Arizona biker Pipes told azcentral.com. “When we tell them we will be there for them, they believe us.”
( Article)
My parents are a part of this organization and they are metal af
They go on runs to protect the child if they feel even the slightest threatened no matter where. If the child needs them to go on vacation with them, they do. Bikers come from across the nation to watch over and take shifts for these kids. And the best part is once you’re adopted into this family as a BACA kid, you’re always one. Even when you’re 40 and the perp gets released from jail, they’ll come meet with you and find your best options for avoiding the person and maintaining the life you’ve built for yourself. Once a BACA child, always a BACA child. In Florida, there’s 100% rate for identifying the perp based on the child’s testimony. Why? Because BACA stands with the child and supports the child so they feel comfortable enough to point out their attacker.
What’s better than a badass biker gang being on your side???
NATIVE AMERICAN CHILD PSYCHOLOGIST WHO IS A BIKER AND NAMED HIMSELF CHIEF HELL YES I’M HERE FOR THAT AND BIKERS BEING BAD ASS TO PROTECT KIDS. HELL YEAH.
Guys? This post changed my life. I saw this post. Forever ago. And thought it was only in america… and wished desperately that they could help me. But then I saw it again, during a bad episode, and checked their site. They aren’t just in the USA
They’re in Canada as well and probably other countries. I met and talked with a native guy who runs the place near me. His name is Shaman. I got in, and I’m considered a BACA child now. Despite being 17, turning 18 when I talked to them. They spent time with me when my abuser was over, they gave me therapy resources. They give you something called a ‘level 1′ where they go to your house with as many bikers as they can, i shit you not a solid 20-40 bikers came from even out of province, and met me. I got to choose my biker name and I got a vest with patches on it and my name on it. They all hugged a Teddybear before giving it to me, and told me if I ever felt the BACA bear was running out of love, to give them a call and they’d refill it for me, and then I got a ride on one of their bikes. Just a day or so ago I went to an annual party with them and they we ate food one of them cooked and had a lot of laughs.
I’ve never felt as loved as I did being a part of the BACA family. They also gave me dog tags with the names, and phone numbers of my 2 workers. So I can call them whenever I feel scared.
BACA is an absolutely wonderful group that will do everything in it’s power to help any child whos been abused.
And it doesn’t end when you’re 18 either. As long as you get in contact/get your level 1 before you’re 18? you’re ALWAYS a BACA kid. I’m 18 now and they still invite me to parties, ask me if I’m okay, and are there for me. They’re still trying to find me resources for therapy.
BACA has changed my fucking life.
I hope you all can read this, and reblog it knowing from someone who fucking been with them, that they are absolutely amazing.
Had seen this before, but never realised that this is on an international level - there’s even a contact address close to where I live (in Germany), very cool (though hoping the only use I’ll ever have to make of it is for donations) ❤
Embossed braille should be standard on computer keyboards.
It would raise braille literacy more than anything else I could imagine - among both the blind and the sighted. Currently braille is actually vanishing due to an increasing reliance on audiobooks and screen readers.
I think that braille has a lot of potential use among non-blind groups. As an alternative to traditional writing for dyslexics. As a way to help photosensitive people type with their eyes closed. Or simply as a means to help sighted people find things without needing the lights on all the time!
Accessibility note: It’s important that braille doesn’t vanish because it’s one of the only written language that works for blind and sight-impaired people. It is necessary for them to interact with the real world where screen readers and audio devices are not available to them, such as elevators, most major metro systems, stairwells, doorways, the bumps in the sidewalk at corners are actually developed in conjunction with audio signals so blind people don’t step off the curb into traffic before the correct time.
Digital technology has made accessibility so much easier for all of us disabled people, but we still *need* the real-world accommodations that we fought and died for
I'm not at all sight impaired (yet, let's be real it's a matter of time for everyone) and honestly having that would be great actually. The F and J keys already have little notches to help you feel where your hands are positioned on the keyboard, why the heck did we stop there? What's stopping keycap manufacturers from doing full braille on each letter?
That wouldn't just help blind people, it'd help people with perfect eyesight type faster and better too because your hands can always know where on the keyboard they are.
We should start describing AI as sorting machines.
LLMs are already often referred to as "word prediction machines because technically it's true, they fundamentally work by predicting what's most likely to be the next word given everything that came before.
But a sentence is pretty much just a list of words, and predicting what the next entry is in a list is basically just sorting a list. So effectively what an LLM does is that it sorts words into sentences and the criteria it sorts by is whatever comes up in training the LLM.
And you can expand that to all sorts of machine learning applications:
- Image recognition? It sorts pixels into words.
- Image generation? Sorts words into shapes and colors.
- Speech recognition? Sorts a waveform into a list of words in order of time.
- Green Screen removal? Sorts greenscreen and non-greenscreen pixels into a transparency mask.
- AI code generator? It sorts words into computer actions.
Anything an AI does becomes immediately funnier when you describe it as a sorting machine.
(and also it becomes very easy and intuitive to understand both the strengths and weaknesses of it, like you can sort words into pixels that are shaped like an image of a magenta moon but you can not sort words into turning the actual physical moon magenta)
We should start describing AI as sorting machines.
LLMs are already often referred to as "word prediction machines because technically it's true, they fundamentally work by predicting what's most likely to be the next word given everything that came before.
But a sentence is pretty much just a list of words, and predicting what the next entry is in a list is basically just sorting a list. So effectively what an LLM does is that it sorts words into sentences and the criteria it sorts by is whatever comes up in training the LLM.
And you can expand that to all sorts of machine learning applications:
- Image recognition? It sorts pixels into words.
- Image generation? Sorts words into shapes and colors.
- Speech recognition? Sorts a waveform into a list of words in order of time.
- Green Screen removal? Sorts greenscreen and non-greenscreen pixels into a transparency mask.
- AI code generator? It sorts words into computer actions.
keep thinking about how I wrote in my dissertation about how every time a new form of public/social space emerges it's immediately popular with kids and teenagers who see it as a chance at freedom and then adults colonise it and kick them out. this happened with malls in the 80s and diners in the 50s and pool halls in the 20s. my dad was doing research on this trend in like 1975. and I was like "yeah so this is going to happen to the internet" and then five years later every government suddenly decided to ban kids from everywhere online. I hate being right especially when I don't even get paid for it
There is definitely a phenomena where people try SO hard to avoid anthropomorphism they end up looping around into this quasi-religious stance that humans have some essential non-biological quality that sets us apart from other animals. Like being so cautious about how you describe emotion experienced by a nonhuman animal that you go "that animal is not 'happy' it's just demonstrating a response to positive stimuli and receiving chemical reward signals" as if that's not also what human emotion is at the fundamental level.
there is a very real tendency of teenagers with anxiety disorders self diagnosing with considerably more stigmatized and impairing mental illnesses (e.g. schizophrenia, DID, personality disorders), but the best response to that isn't to get angry with them for "appropriating" lol. instead you show them coping resources for the problems they're actually having and deemphasize diagnostic categories in general. if an 18 year old is claiming to have alzheimer's, they're probably making an innocent mistake and are in genuine distress. be kind.
Also I think this trend comes, at least in part, from how brushed aside anxiety disorders can be. If your parents and teachers dismiss you with 'oh everyone feels anxious', then inevitably you're going to start thinking that there must be something else going on with you
”You must feel very scared right now; let’s talk about how to help you personally, tailored to your symptoms” will always be more helpful than “stop faking (X) for attention”. If theyre that desperate for attention or an explanation, something is wrong.
romaincrisis
janainaart
origamiaround
gummygunk
areudirty
baratheonpilled
prguitarman
jbbartram-illu
andulkaphoto
