Israel systematically targets the health of pregnant mothers and infants in Gaza. This is a reality every parent in Gaza knows. It has also been confirmed repeatedly by others. Pediatric healthcare professionals in Gaza have been bravely speaking out about how Israel is currently using restrictions of formula into Gaza to torture and kill Palestinian babies.
My friend Nader recently became an uncle to his newborn baby niece, Masa. Due to the traumatic stress and malnutrition experienced under genocidal conditions, many mothers cannot produce breast-milk. Masa, like tens of thousands of other babies in Gaza, relies on formula to survive. Today, Nader told me that this little girl is in the hospital now due to a severe chest infection. She is only 8 days old. Please help her.
I woke up today to the tears of my brother's wife, the mother of little Masa, who is suffering from a severe chest infection and is currently in the hospital under medical care. We implore you, friends, to help us get out of this situation and donate to buy milk and other necessary baby supplies for this child so we can leave Gaza. Please don't hesitate to donate.
Nader's fundraising is vetted (gazavetters #4) and supports his large family of nine people, including small children of which little Masa is the youngest. Israel's systematic targeting of children means that the price of formula, diapers, medicine, and other basic supplies are extremely expensive. Please help him raise the funds needed to help innocent Masa survive genocidal conditions.
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.
#excuse me but are you telling me that the Apollo pic is made with the help of the SUN and the Artemis one with the help of the MOON??? #that's actually so poetic i want to cry
@gorandomshesaid wait i need to sit with this one. wait.
I've recently learned that escaping enshitification is actually really easy. When I was younger I would use FOSS (free and open source software) because I usually couldn't afford anything else. Back then the paid options were usually significantly better than the unpaid options so when I finally got money I started using the paid stuff and completely forgot that Foss existed.
If you don't know why Foss is awesome its not just that its free. Open source software has a huge set of advantages. Most software is proprietary meaning that its code cannot be viewed by other people. Open source softwares code can be viewed by everyone and people can contribute their own code to make it better. Open source is often more secure than proprietary because more people have more eyes on it and more people have contributed to it's security. Its also easy to know how much privacy you have with every software because you can read the code. No more "trust me bro I'm not harvesting and selling your data" you know if they are collecting and selling because the code is publicly available.
In the last two months I have been switching almost completely to Foss. I was worried at first because when I stopped using Foss over 10 years ago the average Foss software was genuinely worse than the paid proprietary alternative. Thankfully things have really come full circle. 90℅ of the software I have tried in the last 2 months works better than the proprietary alternative and is 100℅ less obnoxious.
So here is a list of every Foss software I have tried and recommend. There is way more than this available. This list is just what I have used and like personally. Anyone can feel free to add and we can turn it into a master list. Please just take these as a place to start and do your own research to see if these softwares will work for your use case before you fully ditch your proprietary software.
Operating systems
Graphene os: android alt. Security and privacy focused. The most secure and private smartphone currently available.
Linux mint: easy to use linux, 100℅ better than windows.
Pop!: The Linux distro you should use if you have nividia hardware and want to play games using said hardware. Very intuitive and easy to use.
Kubuntu: Ubuntu Linux with KDE desktop. This is the linux distro one I am currently using and I don't have any plans to jump ship again. Better than windows and better than Mac. The companion app for your phone makes life soooo easy. Its pretty and easy to customize to a ridiculously granular level. No fucking notes.
Kindle jailbreak- ko reader: use jailbreak to free your kindle from the tyranny of the bezos. It will download ko reader which is a Foss OS that has every fucking feature you always wished kindle had and let's you read whatever the fuck you want, and have whatever the screensaver you want (no more ads!). Soooo fuck amazon and use this. I genuinely cannot recommend it enough.
Linux FOSS: (some available on windows and android as well)
Calibre: Foss desktop eBook library. Packed with features. You will want to use this with koreader to make managing your kindle easy.
Manuscript: skrivner alt. Does absolutely everything I need.
Bitwarden: password manager (use a keepass fork if you want self hosted)
Next cloud: private google drive and cloud alternative that you can self host if you want but it's not required. App available
Proton VPN: to the best of my knowledge it is the only Foss no log VPN you can get. You can pay for higher speeds. App available
Quad9dns: free encrypted DNS provider. App available.
News software:
Use any Foss RSS reader and for the love of god stop getting news from social media. Take control of your feed!
Apps (I only know for android)
Fdroid: great app store to find Foss android apps and download them.
Antennapod: you can get all of your podcasts fetched to one feature rich app via RSS feed. No need to rely on spotify or music steaming services.
Openreads: it's good reads but private and 100℅ stored locally. No amazon, no social media aspect, it just tracks your reading, you can import your good reads but if you want to import from storygraph you have to make a good reads burner account, import to good reads, then import to openreads. The menus Navigation on this one is a bit cumbersome but honestly good reads app is worse.
Newpipe: YouTube frontend that let's you have YouTube subscriptions, watch YouTube in the background, and blocks all ads, without logging in to YouTube. You will want to use this one with a VPN set to Canada (or any other country) so YouTube doesn't keep blocking it in order to force you to sign in. But even with that extra step its worth it for the privacy and the lack of ads.
Proton mail: one of 2 more private gmail alts. But you should note that email cannot be 100℅ anonymous or private.
I think that's it. There are still a lot software varieties I am slowly finding.
Since some of you don't seem to understand how this 'new notes' thing works, I'll break it down:
I'm the OP. I'm making this post. If you like, comment, reblog (without comment) on this post, then I'm the one who will see all those notes in my activity page.
However...
If you reblog (with comment), I will get a notification that you did that, but any likes/comments/reblogs (without comment) you get on that reblog will only be shown to you. As OP I won't see them.
If someone adds a reblog (with comment) to your reblog...as OP I won't see that. I won't see any of those notes in my activity page.
Basically, if someone with a large following makes a comment, then they will get all the notes and OP will see nothing. If OP has said something silly because they're, y'know, 21 and it happens, and then someone reblogs it onto the dash of someone with a large following who then dunks on them for fun? OP doesn't see it, doesn't get notes for it, but they're gonna get the harrassment for it in their inbox.
If I, someone with a 5 digit follower count, reblog something to correct misinformation on Ancient Egypt, then OP will never see it unless it was on the original post, but I will continue to get notes on that post even though it's not my post. If I reblog fanart, or just art in general, with a comment like 'Oh this is so lovely!' then OP will not see any of the notes from people reblogging it from me. They'll only see my reblog. So it's possible for an art post by someone else to have 200 notes for them, but 9000 for someone who reblogs it with a comment, and the OP artist will have no idea it's been seen by that many people.
It's killing blow to the community we've built here, by someone higher up who doesn't understand that being able to see all the comments and reblogs is what makes this site the place I keep coming back to.
That's what sucks.
I encourage people to go to tumblr's support page, select contact support, and then in the dropdown menu select 'Feedback' and leave polite and constructive feedback (for those of you who enjoy 'emails worded politely but are a strong 'are you an idiot?', try that way of wording it). They're more likely to listen to you if you're not an asshole about it. I've already gone and done this, and I hope others will too.
And if you love it, Ncuti Gatwa, Ronke Adekoluejo, and many of the creatives involved in this production are nominated at the Stalls to Stage Awards 2026, and you can be there!
Shared here today by Matthew Boroson on Facebook. (ETA: Gaining inspiration from other authors is great. Lifting passages and avoiding giving credit isn’t.)
Tanith Lee was the first woman to win the British Fantasy Award for best novel, for the second book of the Flat Earth series. She died in 2015. You can buy Tales From the Flat Earth here and here .
In 2026, the financial siege that strangled Cuba for more than six decades became a military one.
Five oil tankers seized in a single month. 7.3 million barrels confiscated. One of the tankers was not even under sanctions.
The largest naval deployment in the Caribbean since 1962. US drones surveilling Mexican tanker routes. An executive order threatening tariffs on any country on earth that sells Cuba a single barrel of oil.
Mexico, facing $400 billion in trade exposure to the US, stopped shipments. Venezuela's supply was destroyed by force. No alternative supplier was willing to risk retaliation or seizures.
20-hour daily blackouts. Hospitals on generators running out of diesel. Families cooking with wood.
The Secretary of State testified to Congress that regime change is the objective. The President said: "I think it's just going to fall."
But the siege did not begin in 2026. It began decades ago, and it was never unilateral.
187 nations vote to condemn the US embargo on Cuba every year. 33 consecutive years. The most lopsided vote in UN history.
And every year, every country that votes against it lets its banks enforce it anyway.
The reason is structural.
88% of all global foreign exchange transactions touch the US dollar. 95% of cross-border dollar payments clear through 42 American banks. One country controls the pipes through which the world's money moves.
That is all it takes.
Any foreign bank that processes a Cuba-related payment faces ruin.
BNP Paribas was fined $8.9 billion.
Société Générale, $1.34 billion.
HSBC, $1.9 billion.
Standard Chartered, $1.1 billion.
ING, $619 million.
$13.5 billion in penalties against foreign banks from countries that formally oppose the embargo.
The lesson was received. Most foreign banks now refuse all Cuba operations.
Several countries passed laws making it illegal for their own companies to comply with the US embargo.
Total enforcement of those laws over 30 years: one fine. $15,000. Against a hotel in Mexico City.
The votes against the blockade are symbolic. The fines are real.
And the machinery does not stop at banking.
A US private equity firm buys a Dutch software company. 23 years of Cuban contracts, severed in a week.
A US corporation acquires two Swiss ventilator manufacturers. Deliveries to Cuba stop overnight.
An American cargo company refuses to deliver Jack Ma's donated medical supplies to Cuba. It was the only country in Latin America that did not receive them.
PayPal blocks any transaction containing the word "Cuba." Including orders for a cocktail recipe book.
Cuba does not lose these suppliers to politics.
It loses them to mergers, algorithms, and compliance departments that would rather cut off an entire country than risk a phone call from OFAC.
The result:
35 children on a pediatric ward vomiting 28 to 30 times a day because the anti-nausea drug essential for chemotherapy cannot be sourced from anywhere on earth.
An 89-year-old woman implanted with a pacemaker recycled from a dead patient, two years of battery life, because no manufacturer will sell to Cuba.
69% of necessary medicines unavailable.
Infant mortality rising for the first time in decades.
When one nation controls the infrastructure through which the world trades, and weaponizes that control to deny an island of 11 million people fuel, medicine, food, pacemakers, ventilators, software, insurance, shipping, and banking for more than six decades, while every other nation on earth formally objects and none enforces its objection, the word for that is SIEGE.
if you dont know the context, JK Rowling appears in the Epstein files being invited to the pedophile island and in return, invited Epstien to her Harry Potter play, The Cursed Child. while her full name is redacted, she signs her emails with "Jx" like she often signs her tweets, and Epstien calls her "wordsmith", so we know he was talking to 1. an author 2. someone connected to Harry Potter enough to be able to offer an invite to the play 3. someone incredibly wealthy because those are the only people he communicated with 4. SOMEONE WHO SIGNS OFF AS "Jx"
she brushed it off, calling it "silly", then she illegally deleted YEARS of tracking on her yacht. never any threats to sue despite what would be very clear and obvious defamation and libel case if it weren't true, but none of these scumbags will do it because they know it would all be proven TRUE in discovery
this is a woman who has dumped millions of dollars to make the lives of trans people in the UK miserable. she calls us perverts, abusers, and pedophiles. and here she is, visiting pervert pedophile abuse island to do what they do there. truly every accusation is a confession with these devils
