The summer between the end of high school and the start of college, I wrote a ridiculous play about pirates and put on a staged reading with some friends at an amphitheatre at a local park before a small audience of friends and family. It was never published or staged again. But I just got a message from an old high school friend I haven’t seen in years. He accidentally quoted the play in a conversation with friends, was asked what he was quoting, he couldn’t remember either, and wracked his brain until he finally remembered it was that silly play reading that we did one day in the park over 10 years ago. It made me happy. (The line was, “Huzzah for mercantilism!” by the way.)
A very tiny percentage of creators go on to be famous, but that doesn’t mean that people don’t remember little things you did for years and years. Who came up with most of the world’s most famous jump rope rhymes? Who coined some of the famous idioms we use in daily speech? Who made up ‘Jingle Bells, Batman Smells?” Somehow, all of these things stuck and spread around.
When I was a small child, I saw a high school put on a production of the musical HONK. In one song, the mother duck describes various dangers that her baby should avoid in the water, including fishing line, which could strangle him. A member of the ensemble played the role of fishing line, doing a maniacal laugh and over-the-top strangling motions, and I found it hilarious– and to this day, that’s an example I often think of when talking about how ensemble members can still stand out in theatre. The guy who played the role might not even remember that he did that, but I do.
I took Suzuki violin lessons as a kid. The teacher made up lyrics to some of the songs, and she let her students make some up, too. Now whenever I hear the instrumental of one of those pieces, I always remember these ridiculous lyrics about a skunk that we sang in violin class. I don’t even know which student invented them!
In middle school, I found a video about atoms parodying Bill Nye made by some kids for a school product. It probably had less than 1,000 views, but I think of quotes from that video all the time. They had a parody of “We Will Rock You” with the chorus, “Protons, neutrons, electrons” that I think about a lot.
I just love that this is part of human life. Our memories don’t just pick up quotes from great art, literature, and music, but little things, too.
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.
on the topic of disabled characters, though, i think a LOT of artists give characters mobility aids that their character wouldn't actually benefit from using.
i'm sort of having a mild brain fog moment rn so i can't fully articulate what i'm on about, but i think a lot of able-bodied artists view things like canes and crutches as interchangeable. and they often don't know enough about the wide range of mobility aids to make an informed decision about what their character might actually use based on their disability
i also think that aids like walkers and rollators are sort of viewed as "cringe" on some level, and many people view them to be "just for old people," so many artists will opt for giving a character who has full-body pain and fatigue a cane (which, in many cases, can potentially WORSEN full-body pain) instead. and a lot of people don't realize that most wheelchair users are able to walk, so they don't consider it an option for characters with chronic pain
idk it's kind of frustrating. i've genuinely never, ever seen a character that uses a rollator before. not once. i cannot name a single one. i know of multiple ocs who are described to have full-body pain, dizziness, and fatigue that are drawn as cane users and are never shown with wheelchairs or walkers, though
it's also sort of sad? i've been diagnosed since i was twelve years old, and i only heard of rollators LAST YEAR. and that was only because i saw one irl and immediately asked what it was. i wonder how many disabled people out there don't even know it's an option solely because they're completely underrepresented online and in media
if you're a creative with a disabled oc, PLEASEEE check out this link. it's a very nice and detailed guide for telling which mobility aids suit someone's needs best. it's incredibly important to make the right choice for your character! mobility aids tell you a LOT about someone's lived experience. picking the right one will communicate to your audience #1 what your character goes through #2 that you UNDERSTAND what your character goes through!
"we could give you a link to this mp3 OR we could run it in a proprietary player app that must connect to the internet every time you hit the resume button"
when i upgraded from a flip phone to an iphone and realized i could no longer record and set a custom ringtone because apple wanted me to buy radio pop ringtones, i realized, oh cool new tech isnt made for us it's made to exploit us and we are going to let it happen
EZGIF outperforms just about any other tool for reducing the size of a gif while keeping the quality. very recommend for when you need to upload a gif to a site with filesize restrictions from the 90s (looking at you, tumblr)
The body shape of a fish can tell you a lot about it! It can tell you where they like to hang out, how they hunt, what they eat, and more!
Let's use the Asian Arowana (Scleropages formosus) as an example!
Based off the rounded caudal fin (the "tail"), this fish is not a very fast swimmer
Having the dorsal fin and anal fin right near the end of the fish's body suggests it is an ambush predator
The flat head suggests it hunts at the surface
The upturned, large mouth suggests it catches its prey from below
Arowana are ambush predators that hunt at the surface!
They don't need to go very far very fast, hence the rounded caudal fin (which is the slowest moving fin)
Having posterior dorsal and anal fins are a common trait amongst ambush predator
They primarily hunt at the surface! They will sit near the surface and grab small fish, bugs, crustaceans, and even occasionally a bird!
4. Fish that catch their prey from below sometimes do it via suction. They'll open their large mouth, and as the water gets sucked in, so does anything in the water. And the larger the mouth, more water will more quickly fill into the mouth, catching larger and faster prey
I'm not upset discussions of Sally Ride these days don't leave out the fact that she was a lesbian and therefore the first known queer person in space (albeit a fact only known after her death), but I hate that the fact that what is left out is that she, while part of the Roger's Commission after Challenger exploded, was the whistleblower who made sure the information for the defects of the O-rings made its way to Richard Feynman, who then famously, publicly, and on camera demonstrated how icy coldness (such as the cold and icy weather the morning before Challenger launched) could critically deform the O-rings used and keep them from forming a seal. This was also only revealed after she died. (x)
Whenever Sally Ride comes up these days among my (overwhelmingly queer) friends, we all acknowledge that she was a lesbian and celebrate what a role model she was not just for girls but for queer kids (and adults) too, but everyone is always surprised when I bring up the whistleblowing thing, which I think is damn shame and a disservice to her legacy.
What if water didn't have surface tension and whenever you spilled some, the whole floor of your entire apartment was covered in a 2 micrometer deep puddle
funny how the popular conception of reptiles is that theyre like, slimy and wet. meanwhile they have smooth dry scales, while we mammals are covered in glands that secrete water and oils and other chemicals. where do we sweaty greasy synapsids get off calling lizards slimy
