many things are forbidden by law

There are mentions that some other parts of the city have had more reconstruction since the revolution, but a lot of other districts are described as slums too.

Map Wall - To the east, rolling hillsides: Le Jardin, Stella Maris, the suburbs of Saint-Batiste, swallowed up into the megacity. They sound *rich* to you. This is Revachol East.

You - And west of the river?

Map Wall - Couron. It's somewhere to live. Not bad. Then there's Jamrock -- it's *bad*. People shouldn't live there, but they do. Then Faubourg -- it's almost *as* bad and much larger. Then Coal City. It's the worst.

You - And Martinaise?

Map Wall - It's so small you can't even see it on the map. No... wait. There it is! North of Jamrock, the strip of coast next to the Greater Revachol Industrial Harbour. It looks downright despondent. It's almost Coal City, to be honest.

There are a lot of characters with health issues who can't get proper treatment. In the apartment block, we meet the cleaning lady and Cuno's dad:

Cleaning Lady - "Give me a moment." An elderly woman is leaning on her broom, her knuckles white as bone. She seems to be having difficulty breathing.

Cleaning Lady - "The cold never does any good for my bronchitis..." She sneezes into a dirty handkerchief.

Endurance - This woman's health is failing her. There's not much to do, not in this damp.

You - "Are you alright? Should I call a doctor?"

Cleaning Lady - "I'm fine, fine, don't you worry about me!" She starts coughing, red spots appearing on her cheeks.

Empathy - You're still worried. It's very worrying.

Bronchitis is treatable anywhere with a decent standard of living. This woman works and can't afford to live somewhere less damp that would allow her to recover.

Rosemary in the union of moribund alcoholics establishes that healthcare is a mix of private doctors and substandard charities:

Rosemary - "In a week the goddamned kidneys started giving me all kinds of hell. Finally the missus took me to a private doctor's office -- not a charity, the real thing..."

He only needed a relatively minor procedure, to remove an ampoule that was inserted a week before as an anti-drinking measure.

The charity options seem to involve some support groups for those trying to quit drinking, like the one George attends in Harry's ultraliberal path. Temperance is a popular cause among charities.

Idiot Doom Spiral - He sighs. "It isn't that easy... But I'm working on it. I really am. I even signed up for this evening group in Jamrock."

The options seem to be very limited for anyone needing more help. If Harry hasn't met the Hardies by the time he meets The Pigs, this is how Kim ends up dealing with her:

You - "Should we just arrest her?" (Proceed.)

Kim Kitsuragi - "I don't think there's any need for that. In her current state -- and without the gun -- she isn't really a threat to anyone."

Kim Kitsuragi - "I guess I'll just notify my station. They'll contact the Sanitarium and handle the logistics."

There's a lot of material around about the horrors of real-world asylums, which are infamous for keeping patients too sedated to 'bother anyone' or engage with their surroundings, in institutions rife with neglect and abuse. They don't operate with the same goals as short-term hospital and rehab visits, where people are treated with the intention of helping them recover and function independently again. That sort of hosptial treatment is an expensive private service in Revachol. When discussing what to do with Cuno's dad:

Kim Kitsuragi - "We could take him to Remedie or Saint Batiste, but he doesn't have money for medical services. The Almshouse would turn him down..."

Almshouses in medieval times were typically funded by landowners or religious charities to house a select group of residents, who had to meet the criteria of being established locals who were unable to support themselves, and not being prone to bad habits such as drunkenness or gambling. The elderly and disabled, especially the blind, were prioritised. Rules set by the founders governed the selection of applicants, and they supported relatively small numbers, playing a limited role within wider systems of poor relief.

Countries with modern welfare systems would have other options for dealing with someone like Cuno's dad. Social services could get involved, as an intervention in between the binary of 'doing absolutely nothing' or 'there's an orphanage'.

In real world modern europe, if someone with a drugs problem that had reached a near-fatal stage managed to get themselves to a hospital, they'd get taxpayer-funded treatment even if they weren't able to immediately quit. For some substances, quitting cold turkey can be fatal, and the NHS treatment involves substitution therapy with less dangerous drugs, taking several stages to get people to the point where they have the option of quitting a substance with survivable withdrawal symptoms. Rehab treatments can also involve sedation to manage the risk of withdrawal seizures. With alcohol, there is evidence that the risk of seizures increases as an individual undergoes repeated withdrawals (see https://www.sciencedirect.com/topics/medicine-and-dentistry/withdrawal-seizure ) - so for anyone like Harry who has made previous attempts to get sober, future attempts become more dangerous without medical help. And in Revachol, that help is unaffordable for a lot of people. The support groups are free, but the hospitals aren't.

Noid - "Has anything like this happened to you before? Like... a *seizure*?"

Endurance - It's not your moves that are at fault, per se. You almost certainly had a seizure just now -- and a much more massive one a few days ago, when you lost your memory.

Heart problems also typically need help from doctors, not charities that apparently don't provide the same standard of healthcare as real doctors. It's a fandom meme that Harry can die of a heart attack on day one by failing to get his tie down from the ceiling fan, and even if he passes that check, he stays fragile all week. He can lose HP by kicking things or sitting uncomfortably, and he never has so much HP that he doesn't have to worry about too many of those incidents in a row killing him. In the mercenary tribunal, part of the reason he falls unconscious when shot is because of his heart problems:

Kim Kitsuragi - "Back there, on the plaza? I think you very nearly had a heart attack. Alcoholism has damaged you more than bullets have..." There's a small pause. "We should go."

Harry's gunshot wound became infected, and still is even after Kim treated it to the best of his ability for several days:

Kim Kitsuragi - "I had Garte open the door to your room." He closes the notes. "You were running a low bacterial fever the first night."

-

Kim Kitsuragi - "Reasonably bad. You were shot in the left quadriceps -- that's your thigh. The outer side thankfully, no major arteries were nicked. The bullet was removed and a bacterial infection treated with mercurochrome."

-

You - "Two days of sunshine? I just got a bacterial infection."

Lilienne, the Net Picker - "I'm sad to hear that. Take care of that with ether, will you? Don't get too many RCM men around here. Be sad to lose the first one."

The lack of a welfare state safety net is due to Revachol being an occupied territory without its own government, kept in a 'laissez faire' limbo by the foreign coalition that's occupying it.

Joyce Messier - "The RCM's responsibilities are defined by the Emergency, Wayfarer, and Aliments Acts -- three pieces of legislation keeping the city in a -- let's be honest -- laissez-faire stasis to the benefit of foreign capital."

The cut content 'the collapsing tenement' (see: https://www.tumblr.com/pyrholidon/674198053782175744/the-collapsing-tenement-a-case-file-cut-from-the ) establishes that there's no local government with any responsibility to look after the population, so when an apartment building is about to collapse, the residents officially have nowhere to turn.

Damaged Ledger: But, you see, it's *freezing* outside. And there is no municipal government in Revachol. Nowhere to put these people. Two hundred people can't go to their *aunt's for a couple of days*. And above all -- there is no one to *tell* the building is coming down. No authority but you.

-

Damaged Ledger: Your captain puts them in a half-burned building 10 km South. It's got black mould and no roof, but hey -- it's better than death.

Among the parts of the game that weren't cut, Acele tells us that Revachol doesn't have the tax base that it'd need for public services.

Acele - "Then you really haven't seen much of this *strange world*, because we literally haven't got any income taxes. Just a 20% value-added tax and that's, like... *it*."

The RCM have been trying to fill in for other services where they can, but they're donation funded, and chronically strapped for cash. They handle animal control, look for missing people, and sometimes get called by people with medical queries who presumably can't afford to see a doctor.

Esprit de Corps - Wanna hear what's stupid? Somewhere in a one room apartment on Boogie Street a young man shows Patrol Officer Tillbrook his genital warts, asking if they're *cancer*.

We've seen that some working people can't afford private healthcare in Revachol. There are some whose employers provide it as part of the job, but that's relatively priviledged, being part of what distinguishes Moralintern forces from the people they're oppressing:

Coalition Warship Archer - "Quite good. We enjoy all the standard benefits for Moralintern forces on active duty assignments: Excellent healthcare, foreign service pay, a fully funded pension system..."

Coalition Warship Archer - "... six-weeks paid leave, bilingual childcare, vocational certification credits. Really, it's quite a good career."

The RCM has a station doctor at the 41st, and a nurse at the 57th, so they send Kim to hospital if he's shot instead of treating him themselves. A police force wouldn't be able to operate without treating injuries, but the RCM seems to scrape by with the bare minimum.

Kim Kitsuragi - "The doctors will have to look at this. I hope your station has better medical personnel than 57. This is... a little advanced for a nurse."

When it comes to injuries - in the real world, employers have a legal duty of care to their employees, and one aspect of it is that people are protected from dismissal without being given reasonable time to recover from injuries sustained on the job. (See https://www.national-accident-helpline.co.uk/injury-and-accident-claims/accidents-at-work/duty-of-care ). Either the RCM doesn't have anything similar, or it does but flaunts it without consequence, since they can leave Harry in Martinaise while he's dealing with an infected gunshot wound, without giving him medical leave to recover first. (And without taking him home or to a hospital.) If he's left unemployed, he'd be in the same boat as a lot of other people in the neighborhood in struggling to access care - for the bullet wound, or the seizures, brain damage, heart problems, amnesia, mental health issues, or post-polio issues that are implied by the infant polio and the chronic pain.

In Harry's moralintern quest, Harry can get in contact with the Moralintern committee that has direct responsibility for Revachol, and the conversation can turn to how an interrim period is about to start, where Revacholians will be able to vote for Moralintern-selected candidates to form a government that will still answer to the coalition, as a step towards future full democracy. In the meantime, Harry can complain about the mass poverty and suffering, and the imminent bloodshed by the Krenel mercenaries, and that same committee will say that they won't take responsibility for local issues.

Coalition Warship Archer - "We're sorry, Firewalker. The Committee has declined your request for an emergency address..."

Coalition Warship Archer - "Specifically, they said that 'disputes of purely local concern' do not fall under their purview..."

Coalition Warship Archer - "I... cannot say we agree, speaking for a moment in strictly personal terms. You have convinced us that the situation in Martinaise is untenable, but we are only the second signaller, and there is no appealing the Committee's decision..."

Empathy - Something has cracked in her foundation. Her faith in the Moralintern is no longer as firm as it had been...

Coalition Warship Archer - "They added that they expect the RCM to fulfil its responsibilities 'to a high degree of satisfaction'..."

When the situation is untenable, revolution is imminent. Of course, if an uprising gains traction, things will get worse before they get better.

Evrart is kicking things off with the harbour takeover, but we find out at the end that his plan was to let most of the union die to the mercenaries, let bad press drive them off, and be the one to rebuild after it's all been burned down. He's not going to protect the neighborhood in the meantime - he's been treating everyone as meat shields, like the rich companies that he hates so much.

You - "So what's going to happen now?" (Conclude.)

Evrart Claire - "What was always going to happen. We take the harbour and she fucks off to Ozonne, uncorks a bottle of wine, calls her partners and says they need to distance themselves from this nasty business before the big shit spinner splashes everyone."

Evrart Claire - "Only difference is the Union doesn't have to lose 2,000 men to machine gun fire."

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.