CAFFEINE and REFRIED BEANS

♌ | 26 | Taken

[mamabeckie]

Dear agere blogs, if you're coming to see if I break your DNI:

This is my main blog for info/humor. Tone varies widely between serious and comical, but it's never "adult themed" or graphic. No bigotry allowed here, just jokes and sjw information.

I've curated my dash to show me mostly sfw agere and nostalgia content, which will all be reblogged to my nostalgia agere blog that I won't name here. You'll see it in your reblog notes, but the like notes will be from auntbeckie.

[tadpole-art]

Landscaping

[torstherapy]

[sickerfxck]

Trying somno but we both fall fast asleep expecting the other to do something and end up both having a beautiful nights rest

[charlesoberonn]

I both kinda hate and kinda love that "this peace is what all true warriors strive for" is from fucking Zelda CD-i

[charlesoberonn]

It's unironically such a great response to anybody who venerates war and/or says that modern people are "too soft".

And the full quote starts with "mah boy".

[charlesoberonn]

[pterribledinosaurdrawings]

how it feels to message a friend who's having Problems that you can't do anything to help with.

[warcrimesimulator]

[creekfiend]

its so awkward when people ask me why i dropped out and i have to be like "inadequate disability support" bc no one wants to hear this. they're always like i thought they had to provide that though isn't it the law? girl you might want to sit down i have some bad news about the litigation-based enforcement of the americans with disabilities act

then if i do say that theyre like, couldnt you sue? well theoretically maybe but not without spending more money than i have and putting myself through absolute hell. so no. no i can't.

[mabelsguidetolife]

it’s just this

[anotherapologist]

in tears laughing at this perfect response to a debrief google doc I created to share with friends after they watch The Other Bennet Sister

[linguisticparadox]

Spoke to a gen z person the other night and apparently the young folks don't know about the very legal sites from which you can access public domain media (including Dracula, The Strange Case of Dr. Jekyll and Mr. Hyde, and other Victorian gothic horror stories)?

Like this young person didn't even know about goddamn Gutenberg which is a SHAME. I linked to it and they went "aw yiss time to do a theft" and I was like "I mean yo ho ho and all that, sure, but. you know gutenberg is entirely legal, right?"

Anyway I'm gonna put this in a few Choice Tags (sorry dracula fans I DID mention it though so it's fair game) and then put some Cool Links in a reblog so this post will still show UP in said tags lmao.

[ri-writing]

Spreading the news to my followers - if you weren’t aware of this before, here’s the link to Project Gutenberg - https://www.gutenberg.org/

Project Gutenberg is a gigantic collection of books that are in the public domain.  You can read the books through the site or you can download them in various formats so you can get the format you prefer for your eReader of choice.

It is free. 

It is legal.

I was reviewing the list of the top 100 books downloaded yesterday and I saw a fair few that I had to read for college classes - so if you’re a college student and your professor assigns you to read Plato or any number of older works, check here before you buy a copy.

I reread the Anne series several years back - they were free through this.  I need to reread Pride and Prejudice at least once a year, and my e-book version is from this.  Someone recommended Jekyll and Hyde to me a few weeks back and I got a free copy from this.  When I went to Haworth on my last holiday before the plague times, I brought books by the Bronte sisters with me to read or reread that I downloaded from here.  It’s a great resource.

[linguisticparadox]

Yes yes yes! I was honestly so flabbergasted that this young person hadn't heard of the gutenberg project! It's been around for AGES, maybe longer than the kindle has? And it's such a huge project and wonderful resource! It used to be a household name (or maybe that's just my family, thanks to my dad being a cheapskate nerd [affectionate]). I was so glad to be able to share this resource and others with them though, and I wanted to make sure no one else was missing out!

If you look at the first reblog from me I also recommended a few other resources, most of which were from www.archive.org, home of the Wayback Machine! They run openlibrary.org, where you can check out ebooks of some public domain titles! They even have the Bone series by Jeff Smith!

And archive.org itself has all kinds of public domain media including music and movies! For Dracula fans, here's a radio show adaptation of the book, starring Orson Welles! And here's a 1920 movie adaptation of "The Strange Case of Dr. Jekyll and Mr. Hyde," starring John Barrymore, the grandfather of Drew Barrymore!

I'm so excited to see people falling in love with classic media through Dracula Daily! Let's keep that fire blazing!

[wanderingchaos]

Also, if you can't handle reading things, check out libirvox.org! it's a free audio book project taking public domain works and people doing free audiobooks! there's a lot of great stuff on there, but it takes things in the public domain and makes audio books out of them!

it's a super nice project, and you can find some really nice readers there!

[athenadark]

Also don't think a book is old because it's in the public domain

lots of writers and publishers are prepared to waive future profits for entirely petty reasons

because of this the entire works of Philip K Dick [petty writer who found himself with lots of hangers on during his life] and HP Lovecraft [his publisher - who was his wife and hated him] became public domain on their death

Sherlock Holmes entered public domain this year, it's always worth checking because you can save a fortune

and the more popular the classic - the more likely someone has uploaded it

[the-haiku-bot]

Also don’t think a

book is old because it’s in

the public domain

Beep boop! I look for accidental haiku posts. Sometimes I mess up.

[elfwreck]

Want audiobooks instead?

LibriVox has free public domain audiobooks.

Public domain works in the US are:

Anything published (in the US) from 1927 or earlier (this number goes up every year for quite a while), and

Anything published between 1928 and 1963 that wasn't renewed, and

Anything published before 1989 without a proper copyright notice.

(Don't go looking for things in that third category unless you've studied a LOT about copyright law. Mostly that covers things like "weird little newsletters" and "self-published booklets" and sometimes fanzines. But most publications have a copyright notice in them.)

There's also some oddball exemptions here and there; copyright law is a tentacled mess. But those are the basic guidelines. (Except for audio. Audio has its own set of rules. It's weird.) (I mentioned tentacles, did I not? Double the amount of them you were thinking of.)

There are a lot of works from the 50s and early 60s that were not renewed, especially short stories published in magazines.

Project Gutenberg began in 1971; the first text was the US Declaration of Independence, shared through the university computer system. That was the start of "hey computers + public domain text = FREE BOOKS FOR EVERYONE."

[noswordinourlake]

Adding on that Project Gutenberg is not just Eng language texts either! I know specifically about the French texts because I did independent study French lit in high school and all my sources were Project Gutenberg acquired (Candide my beloathed) but there's many open source texts available in a number of languages.

[stevespookington]

browsing the top 100 books downloaded in the last 30 days can be really fun too, interesting to see how things change

https://www.gutenberg.org/browse/scores/top#books-last30

[rostekhorn]

Oh man, yeah, young people definitely need to learn this. I read so many public domain things when I was fresh out of college and penniless but still needed entertainment. Just going straight to Wikisource works too:

And yes, Sherlock Holmes is in the public domain. But I got bored with Sherlock Holmes after a few months, and became much more pumped when I discovered his mirror opposite, Arsene Lupin. Because when you're not only young and penniless but living through the Great Recession, what you really want to read about isn't the world's greatest detective solving crimes. It's the world's greatest thief robbing fat cats blind while pantsing the police along the way.

[trambrosia]

And you can Ctrl-F find words in electronic texts.

This is so powerful that in the old times they made a whole-ass index of every word in the Bible, called a concordance. It is now possible for every electronic book

[incel-dib]

[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.

[queueniform]

[lostgirl88]

Dónal Finn as James Mori­arty

[lostgirl88]

Dónal Finn as James Mori­arty

[lostgirl88]

Dónal Finn as Thomas Hayward

[soymoss]

starling