As Queen of Hell, I'm taking you down the Evans, Cavill, and Fassy rabbit hole with me... Secondary blog so you won't ever see me follow you. Here is MY WRITING> Stop by. Say hi. I won't bite. Unless I want to. I am the Queen of Hell, after all... All adults welcome. Don't follow if under 18. Just don't.
ok, so, i've reread both what i wrote and also a re-write i did, several times, in fact. i've made some minor adjustments (mentally) and just outlined the chapters in terms of what, generally, is happening in each and who is in the chapter. kinda a little excited
one thing that i want to be really careful about is doing this writing at my computer, which i use for work. i think i have a second laptop that will work, but i need to figure out how/where i'm going to do this.
i mean, i was going to toss (and by toss i mean stick it in my garage) my second desk that i got when my keyboard tray broke. maybe i need to set that up. i'd be able to move my chair easily enough. maybe this is what i do.
Between writing and other stuff I’m doing I’m not necessarily writing every day but today I wrote about 500 words so woohoo! And what I’m finding to be interesting, I’m going back to the way I was writing before which is all the dialogue first and then go back and do all the description and everything else that goes along with it so yeah, it feels like I’m actually getting into writing again.
i know that i wrote recently about how i'm getting better sleep and it's making all the difference in the world and something interesting is happening: i've been getting hungry. i haven't been really hungry in... months? years? can't remember the last time (with the exception of when i have been hungry because i haven't eaten in at least a day). i think that this is a good sign: i think my body is healing, getting to a place where i will have energy. it feels like i'm really healing.
Just mentioned this over at Bluesky: so why not here?
The lady who's the star of this video has won multiple competitions for baking the Schwarzwälder kirschtorte—the Black Forest cherry(-booze) cake—in recent years. She is plainly the queen of this art. Turn on the subtitles (the SWF Handwerkskunst subtitles are very good) and soak up the calm expertise.
...I have to go hunting for the associated links, but you can go to her family's place and do a course in baking this cake. (And don't think I haven't been tempted.) These folks are very serious about the quality of their product. Her husband is a distiller, and they distill their own damn kirschwasser for the cake.
BTW, I did a fair bit of research on the basic recipe back when I was preparing to post it on the old European Cuisines site... and ran into the fascinating info that it is illegal in Germany to call the cake Schwarzwälder kirschtorte if it does not contain the requisite/necessary booze. So if you're ever offered one without kirsch in it, you have an excuse to tease the local baker about that, if you feel so inclined. 😏
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.
Who could have possibly predicted that implicit bias would happen coming from the developers - a not insignificant number being men (don’t know if I can call them significantly white, because I don’t know the population of developers, but I would hazard a guess that a not insignificant number in management/executives are white)?
Who could’ve possibly predicted that companies would prioritize profit over anything else?
Infectious disease, such as COVID, hantavirus, avian flu, or pneumonia
Shot by tech bro who bet on his death date on Kalshi
Poisoned by Vance
Air Force One crashes
Act of God, such as lightning strike
Chokes on Big Mac on live TV
Kristi Noem mistakes him for her dog
Other
Voting ended onMay 31
(anon submission)
* For legal reasons I should state that any polls submitted are not reflective of my views and do not encourage any sort of illegal actions 🤣🤣🤣🔥🔥🔥❤️❤️❤️
I think the thing that annoys me most about AI on a personal, day to day, level is what it has done to grammar checkers. If you've never done a lot of editing, or used to 5+ years ago but haven't really in the last couple years, I can't even begin to describe how fucking BAD this shit has gotten. And as an author it is EXHAUSTING.
I just want to catch spelling errors and accidental double spaces and repeated phrases and whenever I use the wrong too/to or affect/effect and shit. But no. They've shoved AI up the ass of every grammar checking software out there and now they all fucking suck and make the most random, obnoxious, nonsensical suggestions.
And yeah, I can ignore all the times it's trying to get me to cut out any semblance of my own voice, or shove things into the wrong tense, or make the most random suggestions on comma usage. But if it's getting all that WRONG, what is it just straight up missing that I SHOULD be correcting? What real spelling and grammar errors are still lurking in there?
I get why people keep saying this (and other versions of it like "Use Adobe alternatives" and "Use Google product alternatives."). But here's the problem: I do not create in isolation. Even my own 100% personal projects are getting sent to other people whether it's editors or printers or beta readers and unless every single person in that train is using the same products, things can get wonky.
Libre Office and Word handle formatting differently on the back end, which can completely break documents if you move them back and forth between the two. So if I write in Libre Office but my beta readers are still using Word, when I send them a manuscript for review there's a good chance things won't look right and my beta reader will not actually be reviewing what I sent them.
Industry standards are industry standards FOR A REASON. Having everyone on the same workflow can be crucial to getting things done effectively and correctly without creating a lot of extra work. And those things are not going to change overnight, as much as we might want them to.
Yeah, Word, let me just leave this whole chunk of dialogue without the closing quotation marks. That's the thing to do. How dare I have two punctuation marks in a row. It's not like that's how closing quotation marks fucking work.
And you know, for young writers, this has got to be so detrimental just from the perspective of opening your document and seeing a million corrections that, frankly, don't need to be there. If you're a young writer you're likely not going to have the background knowledge to know what is and isn't a good suggestion, you're just going to see a document that makes it look like you made every mistake possible so clearly you must be a terrible, stupid writer and should just give up.
this is what i mean when i say that AI is a race to mediocrity. the AI that is available to the public is really just a large language model and it comes from everything that these companies have scraped off the internet. all those blog posts and comments and tweets and posts on all the various places, everything that you've seen that people have written, absolutely everything that these companies could get their hands on, so of course it is a race to the middle because it takes all that is out there, and that the majority of writing is... average.
(I'm assuming the quantity of quality writing is similar to that of a bell curve:
)
it's not going to be on par with authors, it's going to be on par with your next door neighbor and your parents, and all your family and friends and people you work with. it's taking what most people have done and is suggesting based on that and just... no thank you.
so i know i wrote something a little bit ago that was basically "yeah, sleep rocks!" but the part that i don't know that i conveyed is just how good i feel. i feel hopeful that i'll be able to get the things done that i've wanted to for YEARS. like, seriously.
i took today (friday) off so that i have a 4-day weekend and today has been a good day and i haven't done much of anything. i mean, i have done some things, some that i've wanted to do for a while (many many many months) but just couldn't muster up any desire to do it and i'm doing it and i'm... happy.
i mean, i'm actually hopeful!
i'm going to test some things out this weekend, test some theories that i have, see if they are accurate, because if they are, well, then there's no stopping me! who could have known that sleep was the crux of many things?
Doing a final project in my stats class, we have to pick a subject and collect data on it. We need at least 100 data points, and I figured this blog is big enough that a poll on here could get to that pretty easily!
Doing my project on if it’s more likely to be born in certain months :]
I have gotten the OK from my teacher to collect data using a Tumblr poll, btw. I’m also going to have to send her this post as proof of where I got the data from / proof I didn’t just make up the numbers. So. Behave
It’s really interesting when you finally get consistent sleep. For the better part of, I don’t even begin to know how long, I have been getting very inconsistent sleep. My naturopath gave me some suggestions for some things to take because I really think that sleep is the key for a lot of things and it’s turning out that it is extremely true. Thus far I have taken only one of the supplements she recommended, and I am now consistently getting at least six solid hours of sleep a night. Do I wish for more? Sure, but I’ll start with the six. In the past two weeks the only times that I’ve woken up in the middle of the night was because I actually had to use the bathroom and I was able to get back to sleep very quickly. She gave me two other suggestions and I’ve decided that I’m going to finish the one bottle that I have, start with the next bottle and see how it goes, because it’s entirely possible that it will do a better job. If it doesn’t, however I have something that I can use and I will be using it.
With the consistent sleep, I now am starting to have energy. I am kind of excited about this because lack of energy has been prohibiting me from doing so many things and I’m just very much looking forward to the possibility of my future.
With consistent sleep, I also now have the real desire to truly tackle what it is that I’ve been dealing with in therapy since the beginning of February. it doesn’t feel like these things that have had a hold on me my entire life have a hold on me any longer. I don’t know if it’s actually the case, that will happen through some additional therapy, but it kind of feels like they don’t have a hold on me anymore, even the things that I only in the past couple of months learned about.
I don’t know how long therapy is going to take for me to get to where I want to be, it could be two sessions or it could be 20 years. My hope is that it’s closer to the former than the latter.
I’m also extremely hopeful that I will be able to start writing again. That would make me seriously happy. 
