[femalepope]

My deepest darkest fantasy is that I collapse on the street and I am rushed to the hospital. They perform a bunch of tests and find out I am severely deficient in some kind of vitamin. Then I start taking the vitamin and I become the happiest cleverest person alive because all my problems were caused by this one deficiency

[caracalliope]

Moreover, everyone gathers around to be tremulously compassionate and discreetly admiring: all this time, you lacked the Vitamin? And yet you persevered?

[babyjanedough]

Already know I wanna send this to people on June 1

[spoopyandtired]

ITS TIME

[literalliterature]

[Video description: A short clip from the show Make Some Noise. Erika Ishii, imitating Ebenezer Scrooge, points dramatically and yells, "You, boy! What day is it?" Brennan responds, "It's pride, bitch!" End description.]

[lilstinky-deactivated20221002]

sci fi is about one thing and one thing only.....actors throwing themselves around a room to simulate the ship being hit

[remainsingrey-deactivated202405]

[crystalsoulslayer]

[randomgermancupcake]

[justalurkr]

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.

[watchhisjawdrop]

Paper Link - PDF download

https://s10251.pcdn.co/pdf/2021-bender-parrots.pdf

Digital Library

https://dl.acm.org/doi/10.1145/3442188.3445922

DAIR - Founded by Timnit Gebru

https://dair-institute.org/

[kittysneezes]

I like this article because it’s not what you think it is.

[aprilwitching-deactivated201808]

the world is a strange and wonderous place

[mariopowertennis]

YOU HAVE ENTERED

RADICAL SATURDAY

[cynical-werewolf]

Today’s Friday, though.

[sketchlock]

[ohnoitstbskyen]

Any tips on training media literacy

I just worry I’m missing things others are getting and I wanna actually be able to describe why I like certain things without sounding like a broken record.

I can’t analyze to save my life but I wanna learn

I think the place to start is with the consciousness that all media is a collection of choices. All of it. Everything that is in a piece of media is a decision that someone made, from the big themes to the individual pixels on the screen in a video game, those are only there because someone made a decision that caused them to be there.

With that as the foundation, you can start asking the needful questions.

I find a useful place to start asking questions is "why X rather than Y?"

In any given piece of media you want to analyze, pick a choice that is being made and imagine what it would be like if a different (or opposite) choice had been made.

Say, in Star Wars, what if Luke had decided to join his father and overthrow the Emperor? If the movie had made that choice, how would it have changed the film? What would it feel like is the point the movie is trying to make then? What if Vader was not Luke's father?

Or, what if the good guys' lightsabers were red and the bad guys' sabers were blue? What does that change? What are the things you instinctively associate with those colors, why would swapping them around feel wrong? (or would it actually feel really right? If so, why?)

You can do this for the very big questions, like themes and entire plots, but you can also do it in the very minute. Read a poem and ask yourself why the author picked this word rather than one of its synonyms. Maybe it's for the meter, maybe it's for the tone, maybe it's to match vocabulary. Imagining a different word or a different phrase in the poem, how would that change how it feels to read? What it feels like to hear it spoken out loud?

The answer to many questions you'll ask about a piece of media are entirely banal. Like, in 3D animation, some things look the way they do because that's just how a particular software suite handles that particular task. Light reflections, physics simulation, motion blur, etc. Plenty of video games look the way they do in no small part because they are built within the constraints of their engines. Sometimes paintings have the colors they do not because an artiste is imbuing every choice with profound meaning, but because those are the colors the painter could afford, or which were available at the time. And those banal answers are still useful, even if they don't sound deep.

But the important bit is, the more questions you ask (however banal), the more answers you'll learn, and the more you'll understand about the mediums you're interested in, and the more you'll become able to interpret the choices that are made and work towards building an understanding and constructing a reading.

If you start from the understanding that everything (everything! literally every single thing!) in a piece of media is a choice, a decision that could have been made differently, and you ask yourself why it wasn't, you'll be in a decent position to start to interrogate the art you enjoy.

Outside of that, there are lots of books on art appreciation, and tons of people doing videos on media literacy on YouTube. I think Crash Course has some high-school level courses in English and Theatre and Film on their channel, which can be another good place to start.

Also, just, generally... Media literacy is a skill, not a trait. It's a thing which you practise doing, and in practising it you inevitably become better at it. Nobody starts out good at it, literally everybody is improving with practise all the time, and literally the only way to get any better is to try to do it to the best of your ability as much as possible, working the muscle so it will start to grow.

[the-beard-of-edward-teach]

Hawaiʻi is currently in the midst of a natural disaster if you didnt know

Apparently there isn’t much news coverage of this outside of the islands

Towns are flooded, homes destroyed and collapsed, roads collapsed, lives at risk, gas leaks from the flood damage

Haleiwa and Waialua are currently evacuated because the 120 year old dam is at risk of bursting

Mind you that damn is owned by Dole. Theyve known about it needing to be fixed for years and years and years. Despite having more than enough money they refuse

The state has been trying to buy it out from them for years so they can fix it, but the sale hasn’t gone through

Keep in mind that the Dole family were the ones who illegally imprisoned Queen Liliuʻokalani and illegally overthrew the monarchy.

If I see another goddamn person say how sad this is for the tourists whose “trips were ruined” and compare a messed up vacation to people losing their homes, belongings, and livelihoods, I’m going to lose my mind

I am so lucky that my family or friend’s are safe and the few whose houses flooded didnt have it too bad, but so so so many were not as fortunate

If you haven’t heard anything about this until now, I suggest looking into it

The sirens didn’t go off until the flood had been going on for hours. Our state government is spending so much money on a fucking monorail we don’t need rather than fixing the infrastructure.

It’s been the locals and Kanaka doing the most to help get people to safety from the start

I don’t really know how to end this

I just need to know people are aware

I need to know people are seeing whats happening

[elodieunderglass]

Thousands evacuated as Hawaii faces worst flooding in 20 years

"Don't take this storm lightly," Hawaii Governor Josh Green warned on Saturday, as more rain is expected on Oahu and Maui.

[bardace]

Shut Up, I Don’t Care

[lawbreaker13]

Oh I hate this so much, incredible job OP

[vaporwavesimulator]

hey followers. have you ever wanted to know how it feels to be inside a bag of cornflakes

[officialtokyosan]

ye

[vaporwavesimulator]

enter the cornflakes domain

[runby2]

I fucking hate this website because not only did I click this goddamn link expecting it to be a joke of some sort, but it wasn’t a joke and I sat here spinning the screen around enjoying myself in a stupid bag of cornflakes like the dumbass monkey I am on Tumblr.com, enthralled by being in a bag of corn flakes in

[stealingpotatoes]

i love OCs. like this is my emotional support dressup dolly that i beat the living shit out of

[stealingpotatoes]

actually you know what ELSE i love?? cheerleading my friends while they beat the living shit out of their beloved emotional support dressup dollies

[imissthembutitwasntadisaster]

My mom and I have a thing I call woke college daughter where she'll say something like "men inherently require less work and drama in friendships" and I go "not to be your woke college daughter but perhaps rather than phrasing this in terms of biological essentialism we look at it from the point of view of men being socialized to avoid discussing their emotions which is not necessarily a good thing for either men or women" and by prefacing it with woke college daughter I am acknowledging that even if I'm right I know I'm being annoying about it. And it works.

[aurin-avenir]

“Creative people have trouble recognising their skills as skills, because eventually they feel like second nature. This stuff really is valuable, if it wasn’t, people wouldn’t be stealing it. Creativity doesn’t feel special or unique until you realise people have to plagiarise it.”

— Hbomberguy on plagiarism.

[formerlyluckymanticore]

Reblog if you will never. Ever. Use AI in your writing.

[so-many-ocs]

no dude it's so cool how attached you are to that character who is singled out and ostracized due to the external monstrousness that clashes with their internal spark of humanity. and i love how drawn you are to themes of horror and love, nature versus nurture, otherness, isolation, and the abject. i bet you have normal feelings about your own personhood

[kurgy]

wheres seasons greasons

[kurgy]

[kurgy]

its that time of year again

[theangriestlittleunicorn]

It doesn’t have to be

[kurgy]

its not optional

[muffinrag]

i saw a post about wikipedia losing site visits to slop generators. some of us are on wikipedia like white on rice but i think it'd be fun to encourage everyone to start reading a random wikipedia article every so often. yknow, help out their metrics. learn random shit. click to get a random article!

what kind of article did you get?

a person (or a group like a band)

some sort of animal, bug, bacteria, etc (viruses can go here)

some sort of plant, fungus, whatever the fuck seaweed is,

a location

an historical event

some sort of device, contraption, vehicle, etc

a company, organization, government, etc

any type of game (board, table top, video, etc)

any type of art form (writing, music, theatre, painting, etc)

a scientific article

something that doesn't fall into any of the above categories???

Voting ended onDec 17, 2025

and if you read it, feel free to expound in the tags on what it was about and if you learned anything!

I got an article about yuebeipotamon, "a species of potamid crab from hill streams and pools in north Guangdong, China." 🦀