Nothing reminds me what a goddamn miracle modern medicine is more so than hearing stories about people who contracted the black plague in the 21st century and were prescribed antibiotics for it.
Like yeah man you got the disease that wiped out half of Europe, like, a couple separate times within written history, and we have no clue how many times before that. To cure it you have to take 14 pills and drink lots of juice. You’re gonna feel kind of crummy for a while. It’s vitally important you take all 14 pills.
the thing that blows my mind is blood transfusions. for literally all of human history up until about 100 years ago if you lost enough blood that was it, you were dead, and then people just figured out how to take blood from other people and successfully give it to you and now you can come in to the hospital with a blood pressure of ohfuck/nope, the same color as the linens and they just pop a tube in your arm and casually give you some stuff that another person donated on their lunch break, and you live long enough for the doctors to find and treat your gastric bleed. Insanely cool.
The fact that we can put fluids into people via IV saves more lives than I can actually communicate. There are so, so many more ways to die when we can't do that. You can go from literally at death's door from an illness you have no other cure for, to Basically Fine, You'll Feel Icky A Bit Longer But You're Otherwise Fine and Your Own Immune System Will Work Now, from sterile saline into a vein.
Or even fucking subcutaneous, under your skin. It still gets into your system faster and bypasses any fuckery going on in your gi-tract.
But you want the other end?
I recently got the answer to a crapload of symptoms of mine and it turned out to be Crohn's. Ileal crohn's.
For most of human history there was literally nothing to do about this but hope and pray that your immune system didn't decide to rip ulcers and lesions in your digestive tract to the point where you bled out, or the point where parts of it died and killed you with sepsis, or enough to build up stricture bands of scar tissue sufficiently to cause impactions or any other really gnarly and unpleasant ways you can die because for some reason your body decides the walls of your digestive tract are the enemy and need to be dismantled cell by cell. (Including a fuckload of cancers caused by the constant damage to the cell wall.)
Even as recently as when most of the younger people reading this were small children, mostly all you could do about it was take corticosteroids when you were in a flare. And that was better than Nothing. But at the same time, corticosteroids have a potential laundry list of side effects and you want to take them as little as possible and for as brief a period as possible. And there wasn't a lot else.
I am on a medication with the proprietary name "Skyrizi" and the generic name risankizumab. It's made from taking antibodies from a non-human source and then modifying their protein sequences to be more similar to human antibodies, after which they modify them further in order to make it so that the literal only thing they do is go into my body and bind to something called "tumour necrosis factor" so that this will stop flagging my own goddamn digestive system walls for destruction by the rest of the immune system.
Please feel free to read that paragraph over again.
Modern medicine isn't perfect; there are many things we're just as helpless against as we were in the Days of Eld, and there are many ways its practitioners fail us. But also we can make a thing that goes into my body and says "hey stop self destructing you MORON!" and I have a much better chance than at any other time of not dying young of bowel cancer or bowl impaction! This is fucking insane.
There used to be a common, horrific illness that sailors would get, which was mysteriously cured by limes. People know about this one, it's scurvy. But there are other horrible ways to be sick from vitamin deficiency that weren't considered curable at all, and people had no idea what caused them.
Rickets is a disease caused by vitamin D deficiency where your bones get bendy and grow in the wrong shape (it is most apparent in children). It causes permanent deformity and very easy fractures, along with debilitating pain and persistent dental issues. Historically, it was known that milk, and later, cod liver oil, would improve or prevent it, but the reason was not understood until the vitamin was discovered.
Wernicke-Korsakoff syndrome is a complication of alcoholism that leads to psychosis, dementia, and death if left untreated. Severe alcoholics used to just go completely mad before dying, basically. It ultimately results in permanent memory loss (retrograde amnesia), as well as the inability to form new memories (anterograde amnesia). It is caused by the fact that alcohol prevents the absorption of vitamin B1 (thiamine). It is treatable and preventable by giving the patient thiamine shots - if caught early, before permanent brain damage has occurred, it is fully reversible, although the underlying substance abuse issue still needs to be addressed to prevent recurrence.
Pernicious anemia is caused by vitamin B12 deficiency (in turn ultimately caused by an autoimmune issue causing poor absorption). It causes blood cells to be the wrong size and too few in number, resulting in dizziness and fatigue. It also causes neurological symptoms like tingling in the extremities, poor coordination, confusion, and, in late stages, dementia. There was no cure for pernicious anemia in the past. People would simply become anemic and die from it. That's why it's called "pernicious" - that's an old-fashioned way to say "insidious and deadly," named for its slow onset and then-incurable course. Now it is curable with vitamin tablets or periodic injections.
Cretinism, or, less stigmatizingly, congenital hypothyroidism due to iodine deficiency, is a developmental disorder caused by the inability of the thyroid gland to function properly without sufficient iodine. it causes short stature, intellectual disability, infertility, hair loss, and a large lump in the neck known as a goiter (i.e. a hypertrophic thyroid gland). It was historically associated with poor inland populations living far from the ocean (due to the protective effect of consuming seafood, which is naturally high in iodine). We now simply put iodine in table salt, and this disorder is virtually unheard of in regions where this is the case.
Neural tube defects are a leading cause of birth defects, infant mortality, and stillbirth. The most common nonlethal forms of neural tube defects include spina bifida, hydranencephaly, and encephalocele. These defects are caused by a failure of the embryonic structure that becomes the spinal canal to close properly during development, leading the central nervous system to have a distorted shape that may impair cerebrospinal fluid drainage and put pressure on the brain. In severe cases, e.g. anencephaly, the brain/spine essentially develop outside of the body, which is not compatible with life (anencephalic and iniencephalic babies typically die within hours or days; fetuses with more severe forms are usually stillborn if they are not terminated). The risk of these defects is drastically reduced by taking supplemental folic acid (vitamin B9).
Vitamin K is perhaps the most amazing one on this list. Newborns often have very low vitamin K levels due to the fact that it does not cross the placental barrier easily and is not found in high levels in breast milk. It is only produced by gut bacteria, which babies do not have when they are born, and it takes time for them to acquire the right flora from their environment. Deficiency impairs blood clotting, and in infants, can lead to brain bleeds and sudden, unexplained death. Tiny babies would simply die of brain hemorrhaging for no good reason at all. But if they're given a quick shot of vitamin K at birth, that doesn't happen.
We have cured or prevented so many diseases just with vitamins/minerals.
Researchers are developing a futuristic alternative to LASIK that reshapes the eye without lasers or incisions. Using mild electrical pulses
Cool! You can zap your corneas gelid and then unzap them so they resolidify in a new, better shape! And by colloquial-you I mean “freshly-removed(?) rabbit’s eyeballs” can undergo this thing.
People who are good at stuff are not always good at teaching that stuff to others. Often, in fact, as someone becomes an expert at something, they have trouble viewing the first steps from the metaphorical seventh floor landing.
Try reaching a four-year-old to read “cat” or what 2+2 is and you’ll realize that things that you “just do” are actually a series of difficult steps for a beginner. In much the same way, experts “just do” play an instrument, play a sport, solve an equation, or create art.
Some of the worst teachers I ever had were geniuses, but they had trouble getting into the minds of their students, much like the xkcd comic (“silicate chemistry is second nature to us…”). Teaching is a skill; in fact, it’s a dozen skills stacked on top of each other. Thank your local educator today: the current administration continues to underfund education and the current society underappreciates teaching as a skill (and I haven’t even touched on the classroom management, lesson planning, and social skills required by most educators).
I am not a math person and I don't do rock climbing/bouldering
I am a math person and I don't do rock climbing/bouldering
I am not a math person and I do rock climbing/bouldering
I am a math person and I do rock climbing/bouldering
Voting ended onMay 31
("a math person" is a large term. If you enjoy math in any capacity, it doesn't matter if you are highschool student, a grad student, a hobbyist, a tenure-track professor if they still make those, or anything, as long as you enjoy math, you count!)
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.
Discovering how we can make our research more reliable is always good, and discovering contaminants like this is obviously very important. I do want to drop a few lines from the University of Michigan article linked above before we all go “oh microplastics aren’t a big deal anymore” (because I think that’s an easy conclusion to come to if you just read the headlines):
That’s not to say that there is no microplastics pollution, the U-M researchers are quick to say.
“We may be overestimating microplastics, but there should be none. There’s still a lot out there, and that’s the problem,” said McNeil, senior author of the study and U-M professor of chemistry, macromolecular science and engineering, and the Program in the Environment.
As microplastic researchers looking for microplastics in the environment, “we’re searching for the needle in the haystack, but there really shouldn’t be a needle to begin with,” said Clough, a recent U-M doctoral graduate.
Random thing for people to consider is that since Laika is the saint of one way trips should Felicette be known as the saint of safe landings since she did make it back to the ground safely
tu LANCES félicette ? tu lances son corps comme la fusée ? oh ! oh ! prison pour les scientifiques ! prison pour les scientifiques pendant Un Mille Ans !
“but what if you abort the baby who’ll cure cancer?!” sir the baby who will cure cancer is an organic chemistry major who works at a Home Depot because you use AI to go through your resumes
"I am, somehow, less interested in the weight and convolutions of Einstein’s brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops." - Stephen Jay Gould, The Panda's Thumb: More Reflections in Natural History
one person does not cure cancer. teams of scientists who rely on funding from the government cure cancer. one person's singular novel ingenuity is not what is stopping us from curing cancer; fucking CAPITALISM IS.
Since I hate having to do my own searches to verify stuff, here’s a Science Daily link and the journal article it cites for any similarly lazy-but-conscientious people after me. (And the University of Michigan press release, for what it’s worth.)
I'm instituting a new policy of "if I can't easily read your crusty scanned PDF then I'm sending it back to you, telling you to get your shit together and save your .docx as .pdf, and causing snakes to manifest inside your house"
this but also if you are in accounting and you have an Excel file please do not save it as a PDF or take a screenshot of it and then paste it into another Excel file
Good afternoon to everyone in the notes having a horrible time! Y'all are fighting demons I never knew existed!! I think every person that makes you do stupid time wasting shit like this because they refuse to learn basic computer literacy should be fired!!
"This week I discovered the same pattern, executed by Google. Google Chrome is reaching into users' machines and writing a 4 GB on-device AI model file to disk without asking."
Google Chrome is downloading a 4 GB Gemini Nano model onto users' machines without consent, with no opt-in, no opt-out short of enterprise t
there literally isn't a way you can detect "AI generated text" from regular text. it's just text. I could pretend to write like chatGPT and you wouldn't be able to tell the difference because like. there isn't one. it's not Ontologically Marked With AI Soullessness. Trying to ferret out who is "using AI to write stuff" regardless of what kind of stuff it is, is a losing battle, and one in which people will resort to the academic equivalent of hitting the computer with a blunt instrument to fix it.
Ultimately I don’t disagree but there are patterns you can keep an eye out for, such as:
- strange use of idioms or expressions (e.g. “your question has a large blast radius”)
- overly verbose, lots of filler words and too many transitional, connective-type phrases
- overly-acquiescent, just begging to be useful to a degree that’s pathetic and terrified of coming across as rude even when it’s necessary
- circumlocution, too many abstract concepts or talking “around” something instead of addressing the topic directly
- em-dashes, unfortunately
- no stream-of-conscious or true passion. The “rant of love,” the messy run-on flow of praise that you’ll see here on tumblr, is reduced to a grammatically-correct essay
but those aren't reliable and have massive false-positives that will inevitably discriminate against ESL speakers, autistic writers, etc etc, which was the point of the post. and which is a problem you are contributing to.
stormborns
turnnoffyourmind
yousharknotpass
pixelskylines
titsay
andulkaphoto
2087
baratheonpilled
origamiaround
