Hold on, my train of thought just hit a cow

Just me, 34, German. heroine addict. music junkie. compulsive TV show watcher. ✿

[icantspellthings]

Top 3 things people love insisting they don't have despite it being impossible

Pronouns

An accent

Bias

[sophsun1]

Seinfeld – 7.18: The Friars Club 

[ampervadasz]

[masterkfox]

my two brain cells

[vigilantsycamore]

This is what executive dysfunction looks like

[a-book-of-creatures]

Inside you there are two hamsters…

[lesbianlandry]

shane hollander as kermit the frog

[beholdingslut]

in absolute tears about the pride module at my work

[dikdikpronouncedxylophone]

[bigfoots-biggest-fan]

HOLY SHIT GUYS, I WAS INSPIRED BY THIS POST TO TRY MAKE THE SONG AND YOU WOULD NOT BELIEVE THE SCREAM I SCRUMPT WHEN I DRAGGED THE TRAINING AUDIO OVER THE BACKING TRACK AND IT LINED UP PERFECTLY

Visit Original Link

Tempted to actually put this on spotify so I can secretly stream it at work...

Tagging @batshit-auspol because as an Australian you're the only big account I know who might share (sorry).

[punkrorschach]

happy first day of pride everyone

[enyafans4communism]

fuck, marry, kill: the wound that won’t heal, the past you can’t undo, the ghost that keeps returning

[statistical-distr-of-polls]

FMK - In order

The wound, the past, the ghost

The past, the ghost, the wound

The ghost, the wound, the past

The wound, the ghost, the past

The past, the wound, the ghost

The ghost, the past, the wound

Variation I forgor/bald/vanilla extract/results

Voting ended onMay 17

[astraltrickster]

[emmiegoesinsane]

[mewy101]

I could never find the right way to tell you, have you noticed I've been gone?

[writing-prompt-s]

After bitting a Fae, the Vampire claims that she must serve him, for his bite converted her into his thrall. The Fae claims that the Vampire must serve her, for the vampire ate fae food without her permission. As none of them is willing to give up, they bring the case to you, a lawyer.

[theladyregret]

“Congratulations!”

“What?” They both ask, confused.

“On your new marriage!”

[shaunashipman]

what do you mean that's not what happened

[pmamtraveller]

Luca Ponsato - Does Anyone See My Suffering

[nick-nellson]

PARKS AND RECREATION 2.17 Woman of the Year

[gaynehollander]

One time when the Cens are on a string of away games Ilya looks out the plane window and, out of nowhere asks, “what do you think clouds taste like?” Shane had been deep in strategy mode so he hmms a little before answering, “I mean they’re just water, except it’s like condensed and they would be way too cold to consume.” Ilya nods sagely before saying, “so, like slushy.”

[gaynehollander]

One time, after that, they're at the cottage, sitting at the lake, watching the sunset paint the sky a brilliant crimson. Shane's head is resting on Ilya's shoulder and he's thinking of everything they've been through and how lucky he is that it all led them here. How grateful he is that he gets to share his life with Ilya. He looks up at his husband who looks similarly pensive and asks him what he's thinking, feeling tender and so deeply in love, and Ilya says "these clouds are cherry flavour."

[nettlepixie]

you know what fuck it you can read it on ao3 here but i'm reposting this crackfic in full because it's short and it's not that deep and if nothing else i made myself laugh so maybe it will make you good people of tumblr laugh too

centaurs 🐎​🏒​💪​😤​

Bood: hey @ Luca Haas

Bood: it’s so cute you thought you could just text me direct about this 

Bood: want to share your question with the class 👀

Luca: ☹️

LaPointe: weird choice to ask for sex tips @ Luca Haas but go off ig

Bood: “weird choice”????? 

LaPointe: straight????

Bood: you know what fair

Bood: but let’s focus on the real issue here

Bood: go on @ Luca Haas, ask them what you asked me

Bood: assistant captain’s orders

Ilya: assistant captains cannot give orders

Ilya: more like suggestions

Ilya: requests

Shane: @ Ilya Rozanov I assistant captain order you to hang your laundry out, it’s starting to smell bad.

Holmberg: lmao

Ilya: @ Shane Hollander fine i will do this but because it is sexy husband order NOT because it is assistant captain order

Ilya: now @ Luca Haas

Ilya: tell us what you tried to tell bood in secret

Ilya: real captain’s orders 😉

Luca: FINE

Luca: @ Zane Boodram is it ok if I bring my boyfriend to your BBQ next week?

Luca: He’s visiting from Boston

LaPointe: hold the fuck up

Holmberg: BOYFRIEND

Bood: 👀

Ilya: luca you have a secret boyfriend???

Ilya: @ Shane Hollander and i raised you so well 

Bood: somebody call hayden pike there’s still time to get a fanmail of you and him at the BBQ making out @ Luca Haas

Shane: @ Zane Boodram too soon, dude.

Bood: 🤐

Ilya: shut up about fucking pike i want answers

Ilya: @ Luca Haas what is his name

Ilya: what are his prospects

Ilya: what are his intentions

Ilya: can he fight

LaPointe: “can he fight” lol roz are you trying to steal haas from him or what

Ilya: what

Ilya: no 

Ilya: have you seen @ Shane Hollander

Ilya: why would i do that

Ilya: i mean i will fight him if he hurts you @ Luca Haas

Luca: Thanks?

Young: @ Luca Haas answer the questions we’re nosy

Luca: You can ask him yourself 

Luca: On Thursday

Young: booooo

Bood: i got you at least for question 1 @ Ilya Rozanov 

Bood: haasy dropped the name when he first texted me to ask 

Bood: secret boston boy’s name is cliff

LaPointe: omg lol can you imagine

LaPointe: luca brings his mystery man along next week for the hard launch

LaPointe: and it’s cliff fucking marleau

LaPointe: now that would really be a hollander/rozanov manoeuvre lmao    

* 

Ilya: marly 

Cliff: brooooooo

Cliff: what’s up roz

Cliff: miss u stud 🥵🥵 it’s been too long! 

Ilya: marly

Ilya: i’m only going to ask this once

Ilya: are you having sex with my rookie?

Cliff: you mean luca? 

Cliff: hell yeah brother!!! 

Cliff: luca wanted to wait to tell you in person

Cliff: i think he’s kind of scared of you 🤣

Cliff: i was like relax kid rozy’s my boyyy

Cliff: he’ll get it!!! 

Cliff: in fact bro it’s really all thanks to you honestly

Cliff: and not just because we met at your wedding

Cliff: but i mean i feel like real talk over the years u really opened my eyes

Cliff: like i never would have even thought of it u know

Cliff: but being gay is kind of sick 

Cliff: like a dude just really knows his way around another dude

Cliff: knows how to work with it all if u get me 

Cliff: lol of course u get me i know what u and hollzy are like 🤣🤣🤣

Cliff: but also it’s like

Cliff: when i was hooking up with chicks it was always a bit like damn what’s she really thinking 

Cliff: whereas with luca it’s like

Cliff: i know what you’re thinking bc ur my boyyyy 😍

Cliff: just like ur my boy 

Cliff: my king 👑👑👑

Cliff: seriously man i can’t wait to see u in ottawa next week

Cliff: we should hit the clubbbbbb

Cliff: tho honestly i don't really do that so much since luca locked me down 🔑🔒

Cliff: but u know what i mean it’s like an expression bro

Cliff: like we can hit the club in our hearts

Cliff: but just chill at your place

Cliff: whatever ur feeling dude

Cliff: chill is probably better for me anyway

Cliff: need to save some of that energy for luca if u know what i’m saying 🤣🍆💦

Ilya: marly

Ilya: i’m going to fucking kill you

[fuckrozanov]

if ilya rozanov purred “such a good trick” in my face and then kissed me like i was ancient valuable porcelain with a five figure insurance policy and then proceeded to ghost me for 6 months i would kill him with a hammer

[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.

[thydungeongal]

Also shout-out to the Swedes for just borrowing the French "adieu" into their vocabulary and just spelling it "adjö"

[legov7]

German has borrowed the italian "Ciau", spelling it "Tschau" and only using it as a goodby instead of also a greeting.

[thydungeongal]

I had completely forgotten about this, this is fantastic.

[erlann]

If you enjoyed "ciao" becoming "tschau", you'll definitely want to hear where "tschüss" (German, also meaning "goodbye") comes from!

[thydungeongal]

Borrowed from German Low German tschüß from earlier adjüs, from Dutch adjuus, back-formation from adjuusjes, from French adieu.

We can't let the French keep getting away with this

[ospreyonthemoon]

I mean the french very much aren't getting away with it. Everyone else is taking their language and running off with it cackling with glee. We're all getting away with fucking up french words on purpose

[saxifraga-x-urbium]

#English loves french so much it steals the same fucking word a few centuries apart so it gets two words that have the same meaning#But different spellings -#Like guarantee/warranty#(aside: There was a w/g shift so you can see if the word showed up with the normans or got nicked later#Which is why for e.g. the english call it wales and the french call it pays de galls)#Catch/chase#Gender/genre

[thydungeongal]

I'm so glad someone reblogged these additions while I was away from my puter, I absolutely love doublets (or twinlings as they are apparently called as well!! That's so cute!!)!