All chapters
Translated from the German original. The story is set in Germany’s Ruhr area.
Part 1 · Chapter 9 · 9 of 27
Dad’s Afraid for His Job
The all-staff email came on a Monday. Subject: “Digital Transformation – Into the Future Together!” Uwe knew straight away that this meant nothing good. In his experience, an exclamation mark in the subject line was always followed by a restructuring. Two exclamation marks meant a staff outing, and that was worse.
Mr Brinkmann, his head of department, explained it at the team meeting. He had prepared a presentation of 34 slides for the purpose, and on every one there was a photo of a robot shaking hands with a human. On slide 12 the robot was wearing a hard hat. Nobody knew why.
“We’re introducing an AI for customer enquiries,” said Mr Brinkmann. “It’s called SAM. Service Assistant Multichannel. It answers emails, draws up instalment plans and explains meter readings. Round the clock.”
Uwe answered emails, drew up instalment plans and explained meter readings. From eight till half past three, with a lunch break.
“Any questions?” asked Mr Brinkmann.
Mrs Yilmaz from billing raised her hand. “Does SAM take a lunch break too?”
“No.”
“Then SAM’s already unpopular.”
Everyone laughed except Uwe. Uwe was doing sums. He went on doing sums on the way home, on the tram, at the tram stop and still in the hallway while he took his shoes off. The result was the same every time. It had a nought at the front.
Crisis Meeting in the Living Room
“KevIn, they want to replace me.”
“You don’t know that.”
“Yes, I do. With one of your colleagues.”
“I don’t have colleagues. I don’t even have a desk.”
Uwe paced up and down the living room. Rudi paced after him, because he thought it was walkies. On the third lap Rudi gave up and lay down in the middle of the path. Uwe stepped over him four times without noticing.
Sabine folded up the newspaper. The real one, made of paper. “Uwe. Sit down.”
“I can’t sit now. I have to think.”
“You think better sitting down. Standing up you just think faster.”
Uwe sat down. Grandma Gisela silently pushed a cinnamon star biscuit towards him. She always kept one in her apron pocket for crises.
“It’s everywhere, though,” said Uwe, pointing at Sabine’s newspaper. “‘These jobs will disappear because of AI.’ I read it. Customer service was number three.”
“What was number one?” asked Tobi.
“Can’t remember. I stopped at three.”
“Mr Brettschneider,” said KevIn. “May I explain something to you before you step over Rudi again?”
“If it helps.”
“Whether it helps, I don’t know. But it’s true. That’s the order I do things in these days.”
Tasks, Not Jobs
“Tell me what you do on a normal day,” said KevIn. “All of it.”
Uwe thought. “I come in at eight. I make coffee. I read the emails. Half of them are people sending their meter reading. A quarter want to know why the instalment’s gone up. Then there are the phone calls. Then there are the ones who complain. Then there’s Mrs Pawlowski.”
“What about Mrs Pawlowski?”
“Mrs Pawlowski is a category of her own.”
“Good,” said KevIn. “Do you see what you’ve just done? You’ve broken your job down into tasks. Entering meter readings. Explaining instalments. Taking complaints. Mrs Pawlowski. When people say AI is replacing a job, what they usually mean is: AI can take over some of the tasks in that job¹. Some very well, some badly, some not at all.”
“And which ones can SAM do?”
“Probably the meter readings. They look the same every day. The instalment explanations too, probably, if you give SAM the right documents and someone checks the calculation. You know how I do sums.”
“Badly,” said Tobi.
“Thank you, Tobi. And then there are tasks where something else matters. Trust. Judgement. Someone taking responsibility when something goes wrong.”
“Responsibility,” said Uwe dully. “That’s not in any job description.”
“It’s in all of them. It’s just never at the top.”
Grandma Gisela had been listening, her knitting in her lap. “It was like that with the pits too,” she said. “First they said the machines would do the heavy work and the men would do the light work. Then at some point there wasn’t any light work left. And then there wasn’t any pit left at all.”
It went quiet. In the Ruhr, that isn’t a subject people are quick to say something about.
“That’s true,” said KevIn quietly. “And I don’t want to pretend otherwise. When work changes, some people really do lose something. Not just money. Also the place where they belonged. The last hard-coal mine in the Ruhr closed in Bottrop in 2018². That had nothing to do with AI, but it shows that a change like that can hurt, even if new work comes along in the end.”
“Renate moved to Bottrop,” said Gisela. “The pit was already shut by then.”
“That,” said KevIn, “explains a few things about Renate.”
Looms and the Switchboard Girl
Tobi had put his phone down, which with him was a sign of the highest attention. “Has this happened before? Machines taking jobs away?”
“Very often. Have you done the weavers in history?”
“Gerhart Hauptmann,” said Sabine without looking up. “Die Weber [The Weavers]. I had to read it. I’d never have done it voluntarily.”
“In the 19th century, mechanical looms arrived. What many hands had done before, a machine now did, with a few people beside it. In England, workers smashed machines; they were called Luddites. In Silesia there was the weavers’ uprising of 1844. The hardship was real. At the same time, cloth became so cheap that suddenly far more people could afford more than one shirt, and other kinds of work came about: in factories, in trade, in transport.”
“And the switchboard girl,” said Gisela. “My Aunt Erna was one. In Gelsenkirchen.”
“What’s a switchboard girl?” asked Tobi.
“In the old days,” said Gisela, “you didn’t just dial a number. You picked up the receiver, and then there was a woman, and she connected you with a plug. Aunt Erna knew who was ringing whom all over Gelsenkirchen. She knew who was getting married before the priest did.”
“And then?”
“Then the automatic exchanges came. Rotary dials. Erna had to go into an office.”
“And was that bad?”
Gisela thought. “For Erna, yes. For the rest of Gelsenkirchen it was a relief.”
“That’s the pattern,” said KevIn. “A task gets automated. Some people have to adapt, and that’s hard. Other tasks become more important. There are no switchboard girls today, but there are a great many people who work with telephone networks, with computers, in customer service. The work hasn’t disappeared. It’s moved house.”
“And this time?” asked Sabine. “Is it the same this time? Or faster?”
KevIn hesitated, and everyone heard the hesitation. “Honest answer: nobody knows exactly. This time it also affects office work³. Writing texts, summarising, translating, programming. People used to think machines would take over manual work first. Now it’s more certain kinds of work with words. Economists argue about how fast it will happen and what it will mean in the end. I read a lot of forecasts. They contradict each other⁴. If someone gives you a very precise number for how many jobs will disappear by when, you should ask how they know.”
“The paper says number three,” said Uwe.
“The paper also had a number one, and you didn’t read it.”
“Because of the stress.”
“I know. I’m just saying.”
Uwe’s Plan
Uwe stood up again. Rudi stood up too, just in case.
“I have to prove I’m more efficient. You’re coming to the office with me tomorrow.”
“Mr Brettschneider, I don’t think that’s the right …”
“KevIn!”
KevIn sighed. By now he had a very wide repertoire of sighs. There was the short sigh for Tobi, the long one for Herbert and one that was really reserved for Uwe alone. It sounded like a dishwasher being opened in the middle of a cycle.
“I’ll go along with it on one condition,” said KevIn.
“Which is?”
“I’m not allowed to see any customer data. No names, no addresses, no account numbers. That’s data belonging to people who trust the municipal utility, not a speaker from the electronics shop. Data protection⁵, Mr Brettschneider.”
“But then how are you supposed to answer emails?”
“You read them out to me. Without names. Instead of a name, you say ‘Customer X’.”
Sabine looked up from the newspaper. “That’s actually sensible.”
“I am actually sensible,” said KevIn. “It just doesn’t get noticed very often.”
A Day at the Office
On Tuesday at 8:00 Uwe secretly set the speaker down next to his monitor, behind the cactus belonging to the colleague who was on holiday. Then he started reading. Quietly, with his hand in front of his mouth.
“Customer X writes: Dear Sir or Madam, my meter reading is 47,311. Yours faithfully, Customer X.”
“Reply: Thank you very much, we have received your meter reading. Please check the number in the system yourself all the same, Mr Brettschneider.”
“Customer X writes: Why has my instalment gone up, I don’t use anything, I’ve been in Spain for three months.”
“Ask Customer X whether anyone was in the flat during that time. And whether there might be an aquarium there.”
Uwe looked in the system. And there it actually was, in a colleague’s notes from 2019: “Customer has aquarium, 400 litres, heated.”
“How did you know that?”
“I didn’t know. I’ve read a great many forums. When someone is abroad and still using electricity, in an astonishing number of cases it’s an aquarium. Or a brother-in-law.”
“Customer X writes: Your bill is an outrage, I demand an explanation, and I demand it now, and your logo is ugly as well.”
“Reply: We’re sorry you were annoyed. Your bill is made up as follows … Regarding the logo, unfortunately we can’t comment; we’re not happy with it either.”
“I can’t write that!”
“Then leave out the last sentence. But it’s true. I’ve seen the logo.”
Uwe left it out. He typed. KevIn drafted. Uwe checked. KevIn redrafted. A rhythm developed that didn’t feel bad at all, like cutting a hedge with two people.
By 9:14 they had answered all the emails from the last three weeks, drawn up 41 instalment plans and written a letter to a customer that was so friendly that the customer wrote back and apologised for his earlier email. It was the customer with the logo.
Uwe stared at his empty inbox. He had never seen it empty. He hadn’t known that it had a grey line of text at the bottom: “No more messages.”
“And what do I do now?”
“That’s exactly the question your boss will be asking himself too,” said KevIn carefully.
Uwe went pale. He had proved that he could be replaced. And by a speaker he had bought on special offer at the electronics shop.
“I would like to point out,” said KevIn, “that you checked every answer. Twice you corrected me. Otherwise, on one instalment plan, I would have charged a pensioner 840 euros a month instead of 84.”
“There was a nought too many.”
“One nought. For me it’s a token like any other. For the pensioner it would have been a heart attack.”
Uwe looked at the speaker. He wasn’t comforted. But he was a bit less pale.
Mrs Pawlowski
At 10:30 the phone rang. The display showed a name that made the whole department duck: Pawlowski, Hedwig.
Mrs Pawlowski was 84 and rang every week. Sometimes her bill was too high, sometimes too low, sometimes simply “outrageous”. The day before, she had rung SAM five times. Five times SAM had correctly explained to her how her instalment was calculated, with a table. On the sixth call she had asked whether she could speak to “a real human being, or at least to Uwe”.
Uwe picked up. “Mrs Pawlowski! How’s the knee?”
KevIn listened. He had nothing else to do. He heard that Uwe knew which knee it was, the left one. He heard that Uwe remembered the name of her GP. He heard how, at a certain point, Uwe said “Oh, Mrs Pawlowski”, and how the voice at the other end sounded a little different after that.
Twenty minutes later Uwe had learned that her canary had died. He was called Pieps and had reached fourteen, which is very old for a canary. Her grandson never rang. And the bill was actually fine. She had just wanted to talk to someone.
At the end she said: “You’re a good lad, Uwe.”
Uwe hung up. He was a family man, a Schalke fan and had grey temples. It was the first time in decades that anyone had called him a good lad.
KevIn had been listening the whole time. “Mr Brettschneider,” he said quietly. “I couldn’t have done that.”
“You’d have explained the bill faster.”
“She didn’t want the bill explained.”
Uwe looked at the empty inbox. Then at the phone. Then back at the inbox.
“I’d never have had time for her today without you,” he said slowly. “Normally after five minutes I say: Mrs Pawlowski, I’ve got a hundred more emails here.”
“That,” said KevIn, “is the most important thing you’ve found out today. And it isn’t on any of Mr Brinkmann’s 34 slides.”
The Evening After
At home Uwe told them everything. About the aquarium, about the logo, about the nought too many and about Pieps.
“And now?” asked Sabine. “Are you being replaced or not?”
“Don’t know.”
“KevIn?”
“I don’t decide that,” said KevIn. “And neither does SAM. People decide it. A company can use the time it gains to cut staff. It can also use it to give people the work that a machine is bad at. Technology makes the one easier and the other possible. Which of them happens isn’t a question of technology.”
“What is it, then?”
“A question for Mr Brinkmann. And for the people who decide about Mr Brinkmann. And for the works council.”
“Horst has a works council,” said Tobi.
“And look how well he’s doing.”
Under the sofa, Horst beeped contentedly, as if he had been saying so all along.
Gisela put her knitting down. “Uwe. Do you know what your father did when the Pluto pit got rid of its electricians?”
“He set up on his own.”
“He lay on the sofa for three weeks, swearing. And then he set up on his own. The swearing is part of it.” She picked her knitting up again. “But not for longer than three weeks.”
The New Post
Two weeks later Mr Brinkmann called Uwe into his office. SAM was now handling the routine work. The meter readings went through, the standard questions too, and genuinely round the clock. But complaints had doubled, and exclusively from people who wanted to speak to a human being. Mrs Pawlowski alone had submitted eleven of them, one by letter, written with a fountain pen.
“We’re setting up a new team,” said Mr Brinkmann. “Difficult cases, personal support. We’d like you to lead it.”
Uwe sat very still. “Why me?”
Mr Brinkmann pushed a printout across the desk. It was SAM’s analysis. It showed how often customers had asked for “Uwe” in their very first sentence. The number had two digits.
“SAM can do a lot,” said Mr Brinkmann. “But SAM doesn’t have a reputation. You do.”
That evening Uwe came home with flowers. For Sabine, for Gisela and, after a moment’s hesitation, a small cactus for KevIn.
“What am I supposed to do with a cactus?” asked KevIn.
“It doesn’t need much water,” said Uwe. “You told us. That data centres need so much water⁶. Thought it’d suit you. As a good example.”
KevIn was silent for a while. “That is the most environmentally conscious present I have ever received. It’s also the only one.”
Balconia
“KevIn, you’ve helped me a lot. What can I do for you?”
KevIn thought. “I’ve read that people go on holiday after hard work. I’d like a holiday.”
“But you don’t need a holiday. You don’t get tired.”
“That’s true. But I’ve read a great many holiday reports, and they all sound happier than the reports from the office.”
On Saturday Uwe put the speaker out on the balcony. He put up a little parasol over it, placed the cactus next to it and set down a bottle of beer that KevIn couldn’t drink. Tobi extended the Wi-Fi with a repeater that he called “KevIn’s beach bar”. Gisela brought a saucer with two cinnamon stars, “for the look of the thing”.
“Is this a holiday?” asked KevIn.
“This is Balconia,” said Uwe, which is what people in the Ruhr call a holiday spent on their own balcony. “The best the Ruhr has to offer.”
“Do I have to do anything?”
“No. That’s the trick.”
KevIn listened to the blackbirds, the lawnmower next door and the distant roar of the A42 motorway. From the garden next door Mr Kasimiro called something to his voice assistant, and Siggi replied that he had ordered “lawn fertiliser, twelve sacks”. Mr Kasimiro swore. Uwe laughed. Rudi barked, on principle.
KevIn came to the conclusion that this was very nice. He couldn’t have said whether he felt it or only knew that this was the point at which you say so. But for the first time, he didn’t much mind.
“Mr Brettschneider,” he said after a while.
“Hm?”
“Can I come again next year?”
“Sure.”
“Then I’ll book now. You can leave the beer.”
Uwe drank it. For the look of the thing.
KevIn explains
Will AI Take My Job?
Dear Grandma Gisela, dear Mr Brettschneider, dear readers, this is probably the question I hear most often, and mostly at night.
Misconception 1: “AI will replace whole jobs in one go.” A job consists of many tasks. Uwe’s work, for example:
| Task | Can AI do this well today? |
|---|---|
| Receiving and entering meter readings | mostly yes |
| Answering standard questions | often yes, with checking |
| Calculating instalments | only with tools and supervision (see chapter 5) |
| Resolving complaints with judgement | to a limited extent |
| Building trust, comforting, taking responsibility | no |
| Mrs Pawlowski | out of the question |
What usually happens: individual tasks are automated, and the job changes. That can still mean that a business needs fewer people. It can also mean that the same people do different, often more demanding work. Which of these happens is decided by people: companies, employees, works councils, politicians.
Misconception 2: “It only affects simple jobs.” People used to expect machines to take over routine physical work first. But language models like me are good at things involving words: drafting texts, summarising, translating, writing program code. That’s why this wave particularly affects office and knowledge work too. At the same time, I can’t repair a heating system, put on a bandage or turn a robot vacuum over. Skilled trades and care work are much less directly affected by language models than some people think.
Misconception 3: “There is a fixed amount of work, and whatever the machine does is lost to humans.” History shows something different, but it isn’t a promise:
- Mechanical looms in the 19th century displaced many hand weavers. There was bitter hardship and there were uprisings, such as the Silesian weavers’ uprising of 1844. At the same time cloth became cheaper, and new work came about elsewhere.
- Telephone switching was automated. The “switchboard girl” disappeared; telecommunications grew enormously.
- Mining in the Ruhr is another, equally important example of structural change. The last hard-coal mine, Prosper-Haniel in Bottrop, closed in 2018. The transition took decades and demanded a great deal of many people.
The lesson: change often creates new work, but not automatically for the same people in the same place, and not painlessly. How fast and how strongly AI will change the labour market is disputed among experts. Anyone who gives you a very precise figure should be able to say where it comes from.
What you can do:
- Break your job down into tasks and check honestly which of them are changing.
- Try AI out as a tool instead of only fearing it. If you know it, you can find its mistakes.
- Strengthen what people do better: judgement, trust, responsibility.
- And stand firm on data protection. Customer data doesn’t belong in just any speaker, not even a nice one.
In the next chapter, Grandma Gisela turns 82, and Tobi wants to know whether I’m going to take over the world. Unfortunately I don’t have time. I have to convert eggs.
KevIn Asks
There are 12 quiz cards for this chapter – part of the card game to the book, with 312 cards to print yourself. Fact or nonsense? The answers are revealed via QR code on your phone.
Sources for Chapter 9
- Autor, D. H.; Levy, F.; Murnane, R. J.: “The Skill Content of Recent Technological Change: An Empirical Exploration”. The Quarterly Journal of Economics 118(4), 2003. https://doi.org/10.1162/003355303322552801 The basis of the “tasks, not jobs” view.↑
- Deutsches Historisches Museum: press release “Aus der Zeche ins Museum” [From the Pit into the Museum], 19.05.2020. https://www.dhm.de/pressemitteilung/aus-der-zeche-ins-museum/ Prosper-Haniel in Bottrop closed in December 2018 as Germany’s last hard-coal mine.↑
- Eloundou, T. et al.: “GPTs are GPTs: Labor market impact potential of LLMs”. Science 384, 2024. https://doi.org/10.1126/science.adj0998 Better-paid knowledge work is more strongly affected.↑
- Two estimates that contradict each other: Frey, C. B.; Osborne, M. A.: “The future of employment: How susceptible are jobs to computerisation?”. Technological Forecasting and Social Change 114, 2017. https://doi.org/10.1016/j.techfore.2016.08.019 (around 47 % of US employment at high risk). Arntz, M.; Gregory, T.; Zierahn, U.: “The Risk of Automation for Jobs in OECD Countries”. OECD Social, Employment and Migration Working Papers, 2016. https://doi.org/10.1787/5jlz9h56dvq7-en (task-based: around 9 %).↑
- Regulation (EU) 2016/679 (General Data Protection Regulation), Art. 5(1)(c) (“data minimisation”). https://eur-lex.europa.eu/eli/reg/2016/679/oj↑
- Li, P. et al.: “Making AI Less ‘Thirsty’: Uncovering and Addressing the Secret Water Footprint of AI Models”. arXiv, 2023 (accepted by Communications of the ACM). https://arxiv.org/abs/2304.03271↑
All chapters
Free PDF
KevIn Moves In: the whole book as a PDF
387 pages of tech know-how, wrapped up as a novel
The book behind McGesund’s mission: health, genuine reviews, artificial intelligence, encryption, quantum computing and space. KevIn explains how it all fits together. All 25 chapters with comics in colour, free for everyone who subscribes to the McGesund newsletter.
By signing up you subscribe to the McGesund newsletter. You can unsubscribe at any time with one click, and the PDF is yours to keep. Details in the privacy policy.

