Chat, is it considered “abusive roommate behavior” to release a raccoon into the living space after you have asked your roommate for months to please clean up their messes (they do not pay any of the mortgage)
For context, when I used to live alone I would do something called “Princess Time” where I would do an initial sweep (to remove any significant hazards) and then I would release a raccoon into the living area and clean. This helped because I would 1) feel like a princess and 2) the raccoon would bring attention to things my ADHD brain had decided to ignore and I’d quickly clean that stuff up.
So like, if I’m expected to clean the house now, I will be doing it in the way that is most effective for me. And anything that has not been cleaned up after months of having sit-down talks and sending reminders and being promised things will change, might be deemed “trash” by the trash panda and thrown away.
We haven’t done since we moved into the house, because I didn’t want to cause my roommate or their cats destress or have their things destroyed by a raccoon
I am a raccoon biologist and one of the few people in the state allowed to take in captive bred raccoons that had been possessed illegally. The raccoon in the photos is Moonshine, but she is currently at the animal sanctuary where I work as I had been quarantining multiple new intakes from an abuse case. I still have two males (Rum Tum Tugger and Electra) left in my home enclosure as we are getting them neutered and then hopefully sending them to an AZA accredited zoo.
I wanna make things very clear that underneath all the whimsy, I am a trained professional.
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.
me holding a gun to a mushroom: tell me the name of god you fungal piece of shit
mushroom: can you feel your heart burning? can you feel the struggle within? the fear within me is beyond anything your soul can make. you cannot kill me in a way that matters
me cocking the gun, tears streaming down my face: I’M NOT FUCKING SCARED OF YOU
Spotify Premium ad: “Imagine playing music without interruptions!
Infinite skipping! Replay the song you want! And even do it offline? No ads! Whatever songs you want! For a small monthly payme-”
Me: *nods, turns off Spotify and turns on my MP3 player and does all the things they offer, but for free and with songs they don’t even have*
You can also rip audio files from youtube and find files all over the internet. It is far easier to come across great and lesser known music if you dont limit yourself to spotify.
my archnemesis @baleful-meat-worm graciously entrusted me with drawing its somewhere else jon design. therefore he will be cozy and happy and warm :-) yay
So a couple days ago, some folks braved my long-dormant social media accounts to make sure I’d seen this tweet:
And after getting over my initial (rather emotional) response, I wanted to reply properly, and explain just why that hit me so hard.
So back around twenty years ago, the internet cosplay and costuming scene was very different from today. The older generation of sci-fi convention costumers was made up of experienced, dedicated individuals who had been honing their craft for years. These were people who took masquerade competitions seriously, and earning your journeyman or master costuming badge was an important thing. They had a lot of knowledge, but – here’s the important bit – a lot of them didn’t share it. It’s not just that they weren’t internet-savvy enough to share it, or didn’t have the time to write up tutorials – no, literally if you asked how they did something or what material they used, they would refuse to tell you. Some of them came from professional backgrounds where this knowledge literally was a trade secret, others just wanted to decrease the chances of their rivals in competitions, but for whatever reason it was like getting a door slammed in your face. Now, that’s a generalization – there were definitely some lovely and kind and helpful old-school costumers – but they tended to advise more one-on-one, and the idea of just putting detailed knowledge out there for random strangers to use wasn’t much of a thing. And then what information did get out there was coming from people with the freedom and budget to do things like invest in all the tools and materials to create authentic leather hauberks, or build a vac-form setup to make stormtrooper armor, etc. NOT beginner friendly, is what I’m saying.
Then, around 2000 or so, two particular things happened: anime and manga began to be widely accessible in resulting in a boom in anime conventions and cosplay culture, and a new wave of costume-filled franchises (notably the Star Wars prequels and the Lord of the Rings movies) hit the theatres. What those brought into the convention and costuming arena was a new wave of enthusiastic fans who wanted to make costumes, and though a lot of the anime fans were much younger, some of them, and a lot of the movie franchise fans, were in their 20s and 30s, young enough to use the internet to its (then) full potential, old enough to have autonomy and a little money, and above all, overwhelmingly female. I think that latter is particularly important because that meant they had a lifetime of dealing with gatekeepers under our belts, and we weren’t inclined to deal with yet another one. They looked at the old dragons carefully hoarding their knowledge, keeping out anyone who might be unworthy, or (even worse) competition, and they said NO. If secrets were going to be kept, they were going to figure things out for ourselves, and then they were going to share it with everyone. Those old-school costumers may have done us a favor in the long run, because not knowing those old secrets meant that we had to find new methods, and we were trying – and succeeding with – materials that “serious” costumers would never have considered. I was one of those costumers, but there were many more – I was more on the movie side of things, so JediElfQueen and PadawansGuide immediately spring to mind, but there were so many others, on YahooGroups and Livejournal and our own hand-coded webpages, analyzing and testing and experimenting and swapping ideas and sharing, sharing, sharing.
I’m not saying that to make it sound like we were the noble knights of cosplay, riding in heroically with tutorials for all. I’m saying that a group of people, individually and as a collective, made the conscious decision that sharing was a Good Things that would improve the community as a whole. That wasn’t necessarily an easy decision to make, either. I know I thought long and hard before I posted that tutorial; the reaction I had gotten when I wore that armor to a con told me that I had hit on something new, something that gave me an edge, and if I didn’t share that info I could probably hang on to that edge for a year, or two, or three. And I thought about it, and I was briefly tempted, but again, there were all of these others around me sharing what they knew, and I had seen for myself what I could do when I borrowed and adapted some of their ideas, and I felt the power of what could happen when a group of people came together and gave their creativity to the world.
And it changed the face of costuming. People who had been intimidated by the sci-fi competition circuit suddenly found the confidence to try it themselves, and brought in their own ideas and discoveries. And then the next wave of younger costumers took those ideas and ran, and built on them, and branched out off of them, and the wave after that had their own innovations, and suddenly here we are, with Youtube videos and Tumblr tutorials and Etsy patterns and step-by-step how-to books, and I am just so, so proud.
So yeah, seeing appreciation for a 17-year-old technique I figured out on my dining-room table (and bless it, doesn’t that page just scream “I learned how to code on Geocities!”), and having it embraced as a springboard for newer and better things warms this fandom-old’s heart. This is our legacy, and a legacy the current group of cosplayers is still creating, and it’s a good one.
(Oh, and for anyone wondering: yes, I’m over 40 now, and yes, I’m still making costumes. And that armor is still in great shape after 17 years in a hot attic!)
we don't credit rebecca sugar enough for making the episode with the first gay wedding in a kids show extremely plot relevant so it could not be skipped or cut.
#rebecca sugar has gone on record saying that they knew from the beginning they wanted ruby and sapphire and they put every inch of planning#in to make sure that the studio could not take them out. sugar has said they’d compromised on hundreds of things they’d wanted for steven#so that they had the bargaining power specifically to keep ruby and sapphire’s relationship#and a number of ‘filler’ episodes were created just to establish counter-arguments that might come up when they pitched the wedding episode#the one that comes to mind is the episode about steven and connie getting lost in rose’s room steven’s central conflict about liking their#fave book series’ romantic ending was later weaponised when producer’s were like ‘oh but steven’s a boy he won’t be too interested in them#getting married’ sugar was able to be like ‘no. in this episode it’s established he loves romance and specifically weddings. and in these#episodes it’s shown how much steven cares about ruby and sapphire and their relationship and happiness. you cannot convince me this is not#good and necessary plot development#and they wrapped it up in the season finale and the big climactic point of the diamonds finally coming to attack earth to make the#episode integral to the series no skipping it without confusion. and had ruby wear a wedding dress because international censors took#advantage of her design to give her a masculine va#and sugar made certain that everybody knew This was a queer love story that an entire town supported and admired and that any child watchin#it at home would know they are not alone and that that support is waiting for them out there somewhere#sugar sacrificed the wider story they wanted to tell for that and it was a horrible decision to be given but they made the right choice
I love how the notes for this are just chock full of examples of the most batshit specific things people research for their fanfics. Truly a treasure trove.
romaincrisis
stormborns
gummygunk
turnnoffyourmind
titsay
ellievsbear
andulkaphoto
