Sal's Re-blogging Blog 99% running on a Queue! Tags spoilers.
Art @gramophoneturtle | TWEWY Analysis Posts
Art from my blog will be tagged as '#art blog reblog'. Drawings from before my art blog existed are tagged as '#drawing'.
Avatar is from my art blog here. Header is a Xenoblade X screenshot.
I've said this before on my art blog, but here are my other socials:
Pillowfort & here's a 50 multi-use invite link. If it runs out, let me know!
Bluesky | Mastodon | All Other Socials / Links + Commission Info
My Neocities Website: I've started an art archive and started porting my analysis posts from this blog. Has an RSS Feed so you can get notifications on updates!
Tumblr is the only place where I have separate reblog and art accounts. I might do it on Pillowfort later if it feels right but otherwise it's just one account per platform. Each platform has it's ups and downs and honestly none fully have what I like about Tumblr, but I make do with what I have, as I keep seeing what's happening here.
Black Intersex Woman / Black Interfemme Pride Flag:
Black Intersex Man / Black Intermasc Pride Flag:
These flags are pretty self explanatory! Being a black intersex woman / interfem myself, I'd been wanting to make some pride flags for black intersex folks for a while now, so here they are! 😊😎
MAJOR GAME SPOILERS + bonus photos + yapping under the cut
I made Sissel very early on when I began learning how to sew and make patterns, and while the first version is very special to me, I think this one is a marked improvement. He is shaped a lot better and can also actually kind of sit on your shoulder now! I do hate sewing faux fur though oh my god.
I actually made him all the way back in October, but I never know where to photograph these guys other than in my garden, and it was gross and winter so I just waited until it was nice and green again.
If you were wondering, I do NOT have a full Yomiel cosplay, what you see is all I have. I was wearing shorts + socks with slides lol.
I wanted to take some photos in a different spot of the backyard, but these little baby mourning doves where absolutely camped out there and I am not going to bother them.
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.
ok so, I approached my local library with a proposal to donate a mural as a way to A: build portfolio/gain practical experience and B: give back to a beloved public institution. The director was very enthusiastic about it and i've been working on it since the beginning of March. Come with me as I endeavor to paint what is in all honesty an excessive amount of birds
I wanted the birds to look like they were actually in the space so first thing after doing the draft was to do a lighting study
after that I covered the walls in letters in lieu of a projector/vr headset bc i have neither of those :) Then i take a picture of the section of wall and superimpose the lineart over top of it so I can pencil in the lines
et voila
and that was a whole week on it's own so next comes the paintin' >:)
my chains are broken i am FREE. although i did have a great deal of fun with this, the barring on the wings itself took me like four days and i am READY to move on
this was a week and a half of continuous work so please excuse me for getting a little emotional in the bg 🙏
BIRD NUMBER 10!!! The Male Mallard Duck, Anas platyrhynchos
the male and female ones are gonna be posted separately bc they're taking a lot longer lol but yea! super happy i was able to capture the iridescent green of the head, i found metallic green and blue paint at a craft store that really made his head POP. it looks better in person i promise
ALSO!! As this is the 10th one, BIG announcement. The end is in sight!!!!! I plan to finish within the next 3 weeks and there will be a small dedication ceremony/ unveiling happening at the library to commemorate its completion on the 16th of May. If you live in the Western New York region and want to check it out for yourself shoot me a dm!
Also thank you everyone for your kind words and support throughout this whole process, it's been a genuine treat thinking there are potentially thousands of you out there cheering me on while I paint this 🥹
we're movin right along with bird numero 11!! The lady Mallard!! Anas platyrhyncos
the 16th is looming in the distance so i'm trying to get thru these as quickly as i can so i can have as much time for the GBH as possible. i still need to do the names next to all of them so i've got about a week and a half to finish everything which is GREAT because i have adhd and nothing gets my ass in gear like a fuckin deadline, let me tell you
power couple that they are, here's bird number 12 and 13,
the Northern Cardinals, Cardinalis cardinalis
and NOW that they are complete, ITS GO TIME, in the next five days (library's closed for mother's day 😭😭) i need to have the GBH fully rendered, the names of the birds vectored, weeded, masked, applied to the wall, and then painted, plus additional cattails throughout. I may be able to get away with just getting the GBH done in time for the unveiling and then just have the names and cattails added later, but i'm gonna really try to get it all done in time. BUT, i have a plan. Part of why i take so long on these is because i really am just figuring it out as I do it lmao. there have been many a time where i am sitting on top of the ladder googling "how to paint birds" but I think if i take the time tomorro to do all that figuring out how to approach it beforehand, this will go a lot faster. I may also recruit some of my artist friends to help with the placing of the names... hrmm we'll see.
Anyways, shout out to the librarian who tracked down exactly the thing i needed so i could figure out where to place the highlights in my birds eyes, ur the real mvp
Full Body VGen Commission for LetoLeGaosaure of Jumsoon from Leviathan. The whole piece was so much fun to work with, thank you so much for this request! The gemstone lighting was great to tackle, I think it turned out super well!
Composer doodle from 2024 that I cleaned finally cleaned up a little.
I don't usually put Composers on a white background cause it's harder to make them out but I like it here and probably could have played around with more. Ah well, I wanted something simple to clean up and post today :]
I haven't been on Tumblr much recently but did they remove the option to turn off auto playing gifs on desktop?? I can see the setting on my (out of date) mobile app but I noticed they're auto playing on desktop and it's a nightmare. I tried finding it in the settings but I wonder if I'm just missing it.
Papercraft commission of Sheena and Zelos from Tales of Symphonia! This one was quite elaborate with all the flying costume elements, cards and lightning magic, as well as interesting facial expressions and poses; I had SUCH a blast putting on the game soundtrack and getting down to work!
