A doll youtuber I watch has made a video about the history of Polly Pocket, and she's describing the plots of the dvd specials, one of which features an elderly woman named Ms. Throckmorton, and my reaction was
you may think misogyny is good because it is made up of miso, which is delicious, and gyny, which is woman. and girl miso sounds great. but 👆 it is not girl miso
you may think homicide is good because it is made up of homo, which is gay, and cider, which is delicious. And gay cider sounds great. but 👆 it is not gay cider
“It’s funny,” I told Flewin. “We have an old Nintendo Game Boy floating around the house, and Tetris is the only game we own. My wife will sometimes dig it out to play on airplanes and long car rides. She’s weirdly good at it. She can get 500 or 600 lines, no problem.”
The husband was writing an article on classic video game records, was surprised to find out that holding the Tetris record is a bit of a big deal, and mentions how good his wife is at it.
The guy he’s talking to mentions that the record is 327, way lower than his wifes usual scores of 500-600.
They travel to a tournament, and she goes to do her attempt. Just after she beats 327, and is climbing higher, a judge brings up to the husband that the specific version she’s playing actually has a different record of 545.
She overhears that she needs to beat 500-something, and keeps going, setting the record at 841.
There was a decent but ultimately forgettable fantasy novel I read a long time ago that had a single moment that stuck with me.
The protagonist has just won the world famous sword fighting competition in the big, rich capital and is talking to his mentor, and says something about being the best swordsman in the world. The mentor frowns and tells him that no, he isn't. He is the best swordsman out of the people that could afford to show up to this tournament. There could be a mercenary way out in the mountains, patrolling a snow encrusted fort's walls that could kick his ass and there was no way to know until he was already losing to the guy.
I think about that a lot, and how for every apparently dominant competitor, there might be a fucking ronin out there somewhere capable of destroying them.
Ever since starting to publish romance novels I’ve been checking out the romance books at the thrift store specifically for the clinch covers, as a reference for what I might want to do with my own books.
As a culture we mocked these to extinction but I think we were just afraid of their power. The modern clinch revival still hasn't reached the heady heights of what they were doing in the 80s! The vintage covers can be really quite explicit. These ones in particular were steamy enough they had to be hidden on an inner flap.
This episode of the Smart Bitches Trashy Books podcast where they interview Shirley Green and Sharon Spiak, who were romance novel cover artists in the 80s, is a fascinating look at what a huge industry these covers were. Did you know they had whole photography studios full of props to make these? They’d take photos and turn those over to a painter who’d make something like a couple of these a day. They had it down to a science.
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.
hey if it’s your first time being food insecure because of snap benefit cuts, hi! this ain’t my first rodeo and i’ve got tips.
1. dried beans and rice are infinitely cheaper than the canned/microwave stuff. just wash, soak (in the case of beans!), cook, and you’re done!
2. on that note, beans and rice, when eaten together, make a complete protein that is easily digestible by your body. if it comes down to it, you can live off that alone for a good long while.
3. when shopping, prioritize ingredients over premade meals. a frozen pizza is just one pizza, but flour, tomato sauce, and cheese can make many pizzas.
4. dry pantry goods will get infested if you’re not careful. adding a dried bay leaf to dried beans, rice, flour, etc. works wonders for keeping out weevils and other nasty stuff. adding a few drops of peppermint oil to a spray bottle filled with water will keep mice away if you spray it on your baseboards, too.
5. canned veggies is veggies. canned fruit is fruit. you ain’t too good for it, so get those micronutrients if you can.
6. potatoes are super nutritionally dense and a good source of fiber. use them to stretch meals. just don’t store them next to onions because they’ll sprout way faster.
7. coupon clip like a crazy person, and get ready to use the math skills you swore to leave behind in high school. saving pennies starts with getting the most for your dollar. if all cabbages are $1.00, get the biggest cabbage you can find. if tomato sauce is cheaper by the ounce in a quart jar compared to a pint jar, get the quart if you can. be smart with how you spend.
8. lentils are a great way to stretch any kind of ground meat. they’re really cheap and if you’re like me and can’t cut meat out of your diet completely for health reasons, they help it last a lot longer.
9. portion control is so key. if a meal is meant to make four servings, make it last four servings. it’s better to be a little hungry four days in a row than stuffed for two and starving for two. weigh it out if you have to.
10. never say no to free food. whether it’s potlucks, community events, religious gatherings, or student food pantries, a free meal is one less you have to cook for yourself. pro tip: hit up wednesday night suppers at churches, shabbat dinners at synagogues, and sikh gurdwaras pretty much any time for a free (if not really cheap) meal. you might have to endure some proselytizing, but as long as you’re respectful and mind your business, you’re golden.
Hope you don’t mind me adding some extra tips of my own op!
11. Bullion cubes tend to be real cheap and making things into a soup is a great way to make small portions feel bigger. A serving of beans and rice doesn’t feel like much but when you add a bullion cube and some hot water now suddenly that’s soup and you’ve got a big bowl full + extra flavor
12. If you’re diabetic or insulin resistant and suddenly faced with eating lots of carbs because they’re what’s cheap and available pre cook your food then chill it!! Starches in most everything from potatoes to rice to bread will change form when cooked and then cooled. They start converting to resistant starch and our body basically processes them like fiber instead of like carbs, so you can lower the glycemic index of your rice by quite a bit if you cook it then freeze it overnight before thawing and eating. Is it perfect? No but it helps!
13. Frozen fruit and vegetables can also be a great and less expensive way to get your fruit and veg. Canned is often cheaper but do the price-per-ounce math and check! Grocery stores near me still occasionally do 10 for 10 sales on bags of frozen veg! And if you can get some on sales like that they become a great way to bulk out other foods. Mac n cheese? Add some frozen veg. Hamburger helper? Frozen veg. Top ramen? Frozen veg.
14. Reiterating the never turn down free food if it’s something you can eat. Hit up food pantries, check if Facebook has a buy nothing group for your area, people will give away food for absolutely free! Groups like this are also super useful even if you don’t find food, look for things like toiletries on offer, if you can get your body wash for free that’s a couple bucks you can devote to food.
Dollar Tree, Dollar General, and similar stores are excellent sources of seasonings, herbs, spices, simple canned vegetables and fruits, some frozen foods, and occasionally meats and dairy (at a few of the larger stores).
Also if you live in rural farming areas, see if they'll allow you to glean in the fields after the harvest (especially right now with it being sweet potato season). Often there are still a decent amount of good vegetables left behind on account of smaller size and "imperfections".
Couple things here, for when you do this to people:
1. if you get the “answer my call” text, NEVER ANSWER THE CALL.
They are calling you because they want to have the conversation verbally, and be able to lie later about what they said or didn’t say. Force them to continue via text or email- force them to continue the conversation in writing or not at all.
2. “Lack of 2 weeks notice is unprofessional!” or the other version, “Not providing notice is illegal!”
No it isn’t. Neither is true.
And in the US, all states except Montana are “at will” employment (though you may hear an employer refer to it as “right to work” to make it sound better, it’s the same thing). Sure, at-will employment means they can fire you without cause, BUT! It also means that you are not legally required to give a reason for quitting, or to give notice of any kind.
Is it polite to give notice when you can? Sure. Do bosses expect it? Absolutely. But that does not make you legally required to provide it.
3. The only thing I would change in the worker’s interaction here was their response when initially asked to come in.
Employee: “Hey Mark. Sorry I’m unable to cover the shift tonight because I’m studying for my exam tomorrow.”
Don’t give a reason for your lack of availability. It may be tempting to. You may feel rude if you don’t.
DON’T DO IT.
You do not owe your boss any information about what you do off the clock, and any reason you give will only ever be used against you.
Boss: “Hey I need you to cover Jasper’s shift tonight.”
Employee: “Sorry, I’m not available.”
And leave it at that.
Do not elaborate.
Do not offer additional information.
When you boss asks you to elaborate, because they will, be polite but firm. “With respect, that’s personal. I’m sorry, but I’m unavailable to cover this shift/work late/come in early/etc.”
Be a broken record- you’re unavailable. That’s the only information they need to know, and it’s the only information they have a LEGAL RIGHT to know.
Please stop giving your bosses information they don’t need to know and don’t get to have, because they’re only going to try and use it to fuck you over later.
prguitarman
theartofmadeline
blakekathryn
tangentwithtanner
andulkaphoto
ellievsbear
janainaart
gummygunk
kianamaiart
dailybridgerton
