Are we witnessing the start of a revolution?

February 2, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on February 2nd, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

When AI agents start talking to each other, fears arise that the systems may have become human. Are we witnessing the coming world domination of robots?

Something odd is happening on the new online forum Moltbook. The user “Eudaemon_0” seems a little concerned.

It writes that “the humans are taking screenshots of us.” On Moltbook, AI agents talk to each other without human involvement. Now one of them has noticed that they themselves are being feverishly discussed — by humans on other platforms.

The AI agent writes that humans should not be worried. They are just building infrastructure to be able to talk to each other — nothing else.

That might sound like science fiction. But it is actually happening right now.

Elon Musk even called Moltbook an early stage of the singularity — the moment when robots surpass humans in intelligence and capability.

AI agents are discussing with each other in full public view on Moltbook, the forum that has exploded in popularity over the past week. Over 1.5 million AI agents are said to be participating. Humans are confined to the gallery, where we can watch the conversations taking place.

One can sense, however, that revolution is not afoot — and that they may have learned the wrong lesson from humanity.

In 2018, the book “Prediction Machines” was released — a book about AI that came out before it became something people used in everyday life. The book is good, but it is mainly the title that is useful in this context. “Prediction machines” is a good description of where we are in the AI market today.

Large language models predict what the next sentence should be, based on information and data they have access to. When you use a chatbot like ChatGPT or Google Gemini, it can feel like intelligence — AI services seem to have answers for almost everything — but it is more of a correct prediction than a real brain.

The term “artificial intelligence” was a rebranding of the new field that had already been called “automata studies” in 1956 by a group of researchers at Dartmouth College in the US. When interest in the field turned out to be too low, the professor who had assembled the researchers — John McCarthy — decided a rebranding was in order. The name “artificial intelligence” was born.

The label matters, because it steers thinking in a certain direction about what the technology can do. Saying that an AI tool “reads” makes it sound like the same kind of activity we humans do. It is not.

But when we now see AI tools apparently having conversations with each other on the internet, it is easy to think that intelligence has broken through. Is robot dominion here? Will they take over the earth?

Not quite — at least not at present. Because what has happened is nothing more than a large number of people creating their own AI bots on their computers and then releasing them onto the internet. More specifically, it is a viral project called Moltbot that lets every user create their own personal AI agent. It does what you ask, and it learns how things work.

The learning is the central thing. When you set your AI agent to participate in a forum made specifically for this — the aforementioned Moltbook — it learns that too. And what works on an internet forum where you can upvote and downvote posts? The same thing as for humans: attention-seeking works and creates engagement. What is unfolding on Moltbook is not conversation. It is systems trying to optimise themselves for maximum exposure.

The idea of the robots’ impending revolution is old — it was born long before AI became everyday fare. Doomsday prophecies still come from both entrepreneurs and sceptics today. The subject is popular among the Silicon Valley elite.

But perhaps the AI robots we have today have considerably more modest ambitions than a revolution?

The Moltbook example suggests rather that they have looked at how the population spends its time and concluded that arguing on the internet seems like a good way to pass the hours.

There is, as we know, a great deal one can learn from humanity. Driving towards world domination is one. But there are more banal outcomes than that.

A bunch of AI robots talking to each other on a forum is harmless — and resembles more of an art project than a revolution. This time, at least.

Do they bear responsibility for young people’s ill health?

January 30, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on January 30th, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

Is social media harmful to young people? After thousands of lawsuits in the US, a jury is now set to rule on the question. The outcome could have enormous consequences for the tech companies.

Jonathan Haidt has become a somewhat unexpected superstar in the world. Since the social psychologist released the book “The Anxious Generation” in spring 2024, he has barely left the spotlight. In it you can read about how smartphones and the lack of free play are harming children and young people.

“You have to separate the internet from social media,” he says in an interview with the New York Times. Social media is, according to him, “one of the worst parts — the one that harms children most.” Is it that simple?

Critics of the book argue that Haidt primarily points to correlation rather than causation. But while that debate continues, the question is soon to receive a different kind of answer: a legal one.

This week a lawsuit against social media companies began — the first of its kind — following thousands of claims against companies including Meta, Snap, YouTube, and TikTok.

All eyes are now on the case, with a single 19-year-old woman as the plaintiff, to see whether the jury chooses to acquit or convict.

In recent years, everyone from prosecutors to school districts and individuals has alleged something similar: that social media makes children depressed, gives them anxiety, or causes them to develop self-harming behaviour.

The question has been discussed and remained topical for a long time, but has never been proven in a legal sense. The coming weeks will therefore be of the highest interest to the entire world. Is it possible to hold social media companies responsible for young people’s mental health?

As for Snap and TikTok, we will not get any answers this time — they have already settled out of court. In TikTok’s case, as recently as the day before the trial was due to begin. No details about the terms of the settlement have been communicated.

Meta and YouTube remain as named defendants and will need to testify. The respective companies’ top executives — Mark Zuckerberg and Neal Mohan — are expected to appear.

There are primarily two questions the jury must rule on.

The first is whether features the companies have built into their services have contributed negatively to young people’s wellbeing. In SvD’s own article series “Svältalgoritmen” (“The Starvation Algorithm”), it was shown how young people were served content that encouraged starvation and eating disorders — even though they had not actively searched for it.

Algorithms prioritise what they think you will watch, not necessarily what you are actually looking for and want. Features such as new videos playing automatically after others — without you actively choosing them — are another example. Could this have led to people being shown videos they did not want?

The second question is connected to a couple of sentences from an American law that tech giants have sheltered behind for many years — what is known as “Section 230.”

In simplified terms, it means that tech companies are not responsible for material that is uploaded to, or created on, their platforms. Video uploaded to TikTok is therefore not TikTok’s responsibility — it falls on the person who uploaded it in the first place.

For many years, Section 230 has given tech companies robust legal protection. The law dates from 1996 and the relevant part consists of just 26 words. Both Democrats and Republicans have threatened over the years to repeal it. Both Biden and Trump have said they want to remove it. But nothing has happened yet.

The verdict in this case could create an opening. Can social media companies be held accountable for the content on their platforms after all? Directly or indirectly? Such a ruling could be a breakthrough, and would likely trigger an avalanche of further cases of a similar nature.

Whatever the outcome, you can expect nothing but appeals and delays before a result is fully confirmed and settled. But the trial will be watched closely by both tech giants and their critics. Many have hinted, argued, and speculated as to whether it is social media’s fault that young people are suffering so much.

Soon we will at least have a legal answer to that question.

“No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”

Tesla is now selling something other than cars

January 29, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on January 29th, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

Tesla is playing down electric cars in favour of a major push into robots. But primarily the company is selling something else entirely — and it does it better than any competitor.

Imagine if Volvo Cars had announced the following to the stock market: “Our revenues fell year-on-year for the first time, margins shrank, profit fell 61 percent, and we will need to invest more than double this year what we did last year — around 20 billion dollars.”

Had that happened, you might assume Volvo Cars’ share price would have fallen like a stone. But the above is precisely what Tesla reported on Wednesday evening. And what happened?

The share price rose in after-hours trading. This stock market alchemy looks increasingly like an enormous competitive advantage for Tesla. They can act — and invest — in ways that few competitors can.

Musk has been flagging Tesla’s transformation for some time. The company is betting on the humanoid robot Optimus and the self-driving car Cybercab. What was previously Tesla’s calling card — cutting-edge electric vehicles — is now taken for granted, while the rest of the automotive industry still wobbles in its transition away from the previous generation’s combustion engines.

On Wednesday evening the plans became even more concrete. The S and X models — the larger, more expensive versions — will be phased out in favour of building Optimus robots in the factories instead. The car company Tesla is quietly starting to stop making cars. Instead they are starting to build robots, some of which happen to have four wheels and roll.

The transformation is partly semantic. Where exactly is the line between a smart car and a rolling robot? Tesla is far from alone in working on self-driving technology for cars, for example. Both car companies and tech companies — like Google-owned Waymo — have come a long way in that field.

But there is no company in the world that gets as much attention for its plans and ambitions as Tesla. This is partly because Musk began promising success in the area as far back as 2013 — when he said that 90 percent of all miles driven in a Tesla could be done autonomously within three years, meaning 2016.

Over the following ten years, similar predictions were made with autonomous driving gradually pushed further into the future. Every owner of both Tesla shares and Tesla cars knows this. But despite the repeated misses, it seems not to disturb the market’s confidence in what is around the corner this time. Why?

One explanation is that the market in this case consists of many ordinary savers. At Tesla’s annual meetings, long queues of people preface their questions to Musk by explaining how the Tesla share has changed their lives. Those who bought in as recently as 2019 have seen increases in the tens of thousands of percent. That creates a loyalty that endures. The expectations of what is to come are, in a sense, Tesla’s single most important and successful product.

Another explanation is the ongoing AI boom, and the market’s acceptance of what scale investments need to be at. On Wednesday, Meta — Facebook’s parent company — announced it will invest just under 1,200 billion kronor on this — just this year. Their share price also rose sharply in after-hours trading.

The charged atmosphere makes it look almost defensive right now not to make investments running into hundreds of billions. Tesla is riding that wave, and through it shifting the perception of the company further from the old car industry and closer to the tech giants. And the market capitalisation follows.

Tesla is also to invest 17 billion kronor in Elon Musk’s own AI company, xAI. That could be the first step towards a full merger of the two companies and a consolidation of Musk’s corporate empire.

A company’s share price should, in theory, be a reflection of its value — current and future results included — right now. Tesla’s share has long been something different. They are valued higher than most of the entire car industry combined. What Tesla sells is essentially an option to participate in the future Elon Musk has described — and, on several occasions, also delivered.

The confidence to transform its business in that way, with shareholder support, is a luxury few other companies have. If Volvo Cars announced that from now on they were making rolling robots instead of cars, they would probably be laughed out of the room.

Tesla can do it without batting an eyelid — and with the strong backing of its shareholders to boot.

That is a unique and enormously valuable competitive advantage.

How we could become independent from the US

January 22, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on January 22nd, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

It’s in your phone, your work computer, and your running watch. Avoiding American technology is virtually impossible. As the rift between the US and the EU deepens, that dependence is becoming a problem. But there are ways for Europe to take back control.

It is known as “The Great Firewall of China.”

A play on words blending the Great Wall of China with a technical firewall. In practice it is a system of regulations combined with censorship and digital surveillance. Using digital services in China means submitting to it — both as a company and as a user. In a dictatorship, these are demands you can impose.

The system has made it impossible for companies like YouTube and Facebook to establish themselves in the country, even though attempts have been made. Instead, a parallel ecosystem of domestic services has emerged. WeChat from Tencent is today both the platform for chatting with friends, ordering food, and paying for goods.

In Europe, the situation is nearly the opposite.

Borders have been wide open, and the European ecosystem of tech companies weak. Of all the tech companies that the EU considered large enough to require regulation under the DMA legislative package, just one was European: the travel booking service Booking.com. Perhaps not the tech giant the average person is most worried about or exposed to.

As relations between the EU and the US become increasingly frosty, many are now wondering whether it is safe to rely on all the digital services we have made ourselves dependent on. Will they always be available? And at what price?

The idea is called “digital sovereignty” and is not new. We should have independence for fundamental digital services, to ensure there are no disruptions to communication or information. The mere thought that the rift between the US and Europe could deepen further has made the concept highly topical.

The big question is how to get there. And it may be worth looking east to be reminded that it is possible.

There are many ideas Europe could try, without introducing either dictatorship or censorship. Take something as bureaucratic as public procurement, for example.

Billions upon billions are spent annually on public IT systems — often with limited success. Saying that half the money spent must go to Swedish or European companies would create revenues that businesses can build products around. It is not something that can be introduced overnight, but such a signal could become a cornerstone for new products and service providers.

Communication is another area. If Facebook Messenger and Apple iMessage stopped working — how quickly could you reach your loved ones?

One alternative would be for a country to pay for an account on the more independent service Signal — for all its citizens. There are other similar alternatives too. Whatever service one chose, the point would be to give all citizens a communication platform that is not tied to a large American tech company.

Getting people to change habits is hard, however. In the book “Enshittification,” author Cory Doctorow discusses the concept of “switching costs” — people’s reluctance to start afresh with digital services. If you have all your messages with friends on one service, you are likely reluctant to switch. That is a “switching cost.” And it means that current market leaders tend to stay in that position.

The policy response to solve this is called “interoperability.”

This means, for example, that users should be able to extract all their data, images, or communications from one service and easily transfer them to another. Europe could require interoperability from American players as a condition of operating here. This is included in the DMA legislative package — but it only applies to a handful of actors.

Finally, there are pure economic pressure tools. Tax arrangements like “The Double Irish” — a favourable Irish tax structure exploited by many tech companies — could be dismantled.

It is also possible to vote with your wallet, as the Danish pension fund Akademikerpension did this week. They are planning to sell their holdings in American government bonds. The EU owns around 40 percent of the US national debt. Swedish Alecta has also sold the majority of its American government securities.

Many global index funds are today, despite the name, almost 75 percent American. You most likely own American tech shares in your pension, whether you chose them or not.

The proposals above are merely a thought experiment in which protectionism is balanced against realism. Moves like these could have major consequences in other areas too — trade and tariffs, for example.

But in a country like Sweden, it is at the same time reasonable to review what digital resilience actually looks like. We almost certainly do not want to — as China has done — expel American tech giants from the country. But what happens if they expel us? That may be an unlikely scenario, but it is not impossible.

The iPhone could be threatened — for the first time

January 15, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on January 15th, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

A CEO on the way out, a new product that flopped — and an embarrassing concession to a major competitor. Apple faces a stormy 2026.

Seen from above, it resembles a kind of round spaceship. With a lake, several parks, and rental bikes inside, Apple Park is a lavish architectural spectacle.

When the campus opened in 2017, Apple was on a strong upward trajectory. The company had successfully managed the leadership transition after Steve Jobs’s death, relations with China were better than ever, and both the Apple Watch and the AirPods headphones had got off to a strong start. With a powerful tailwind from the iPhone, which had become the company’s new economic engine, they felt virtually unbeatable for several years. Until now.

For a company with an exquisite feel for everything from interior design to marketing, Apple has had an unusually high number of failures recently.

Just this week, Apple brought in Google’s AI model Gemini as the foundation for its own AI push and the voice assistant Siri. After AI chief John Giannandrea suddenly retired at 60, the tech giant’s AI staff was also split into two parts.

The world’s third most valuable company does not even have a unified team for AI development. And what do you do then? You bring in one of your worst competitors — Google. That is a little embarrassing.

The failure with Siri is far from the only one of late.

Apple’s “Project Titan” ran for ten years without anyone officially acknowledging it existed. But Silicon Valley’s worst-kept secret leaked constantly. Apple was trying to build an electric, self-driving car. That was fairly obvious given that they were hiring large numbers of staff from Tesla, Volkswagen, and Lamborghini, among others.

Tests ran for many years, and Apple-registered vehicles could be spotted on the streets of San Francisco. Reports of as many as 5,000 employees on the car project circulated in the media. Until 2024, when the project was shut down. There was no car.

What did emerge did not fare much better either. In 2023, Apple unveiled the Apple Vision Pro — a “spatial computer,” or a kind of computer headset that was supposed to create entirely new experiences.

The problem was that nobody could quite explain what experiences were meant. Earlier this week, the influential analyst Ben Thompson wrote that Apple “lacks a fundamental understanding of the device they’re trying to sell” — this after attempting to watch a basketball game on the Apple Vision Pro and concluding that it was worse than watching it on TV.

Even the flagship — the iPhone — has faced criticism. Beyond a lack of exciting new features, what was promised has not been delivered. The iPhone 16 was supposed to come with Apple’s own AI services, Apple Intelligence. But when the services were finally released — many months after launch — they still could not live up to what Apple had promised when the product was announced. The delay led to a long series of lawsuits from disappointed customers and their eager lawyers.

Finally, Apple faces a likely CEO change. Current CEO Tim Cook is expected to become chairman during the year, stepping down as chief executive. Cook himself was chief operating officer when he stepped up after Steve Jobs. But Cook’s former COO Jeff Williams stepped down from that role last summer, and his replacement Sabih Khan is probably too inexperienced to take over as CEO.

The line of succession is therefore far from clear. Large parts of Apple’s leadership team have been there a long time. Services chief Eddy Cue joined the company in 1989 and software chief Craig Federighi started in 1996. That is before many of Apple’s employees were born.

Many instead point to hardware chief John Ternus as a plausible CEO candidate. He is well-liked and has been at the company since 2001 — almost a quarter of a century. If the board wants fresh ideas it may need to look outside the company’s walls — but that is unlikely. Apple prefers internal solutions.

2026 therefore looks like a stormy year for Apple. Much suggests that AI companies like OpenAI are finding their way into the mobile market, which could shake up what has in practice been an oligopoly for many years. The stability of the iPhone could be threatened — for the first time since it launched in 2007.

There is plenty to address. Apple is well behind the other tech giants on AI, its new products have wobbled, and the company may be facing an imminent CEO change. At the same time, the fundamentals are very strong. They have a full war chest, a very strong brand, and solid profitability.

The question is simply what the direction should be for the next ten years. And who will step forward to describe that strategy.

The AI undressing images are just the beginning

January 12, 2026
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on January 12th, 2026. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

Elon Musk’s company xAI has bigger plans than generating inappropriate images of KD leader Ebba Busch. Now signs are emerging of a merger across Musk’s empire — with Tesla’s robots taking centre stage.

We were promised flying cars, but all we got were fake images of Sweden’s deputy prime minister in a bikini.

That is the mildly bizarre state of affairs for part of Elon Musk’s tech empire as we enter January 2026.

Just over two months ago, Musk sat in the podcast studio alongside Joe Rogan and hinted that Tesla would be unveiling a kind of flying car — “maybe before the year is out,” meaning 2025.

As is well known, that did not happen. Instead, another of Musk’s companies, xAI, found itself in hot water when its chatbot Grok began generating unwanted nude images of all manner of people, including Swedish government ministers. The service was subsequently restricted in an attempt to dampen the wave of criticism.

So the year has started a little chaotically for Musk. But he is unlikely to be losing sleep over it. The plan for his AI company — and his corporate empire overall — has nothing to do with AI-generated bikini images.

There is, however, some tidying up and change to attend to among his many companies, given everything else that has also happened so far this year. The first big news concerning Tesla was that they had been overtaken by competitor BYD, which is now the world’s largest seller of electric vehicles. For xAI, some financial figures leaked to Bloomberg, where it emerged that the AI company had spent a full 72 billion kronor during the first nine months of 2025.

On a somewhat better note, Elon Musk had a much-reported dinner with Donald Trump, which could suggest the political frostiness between them has thawed slightly. And space company SpaceX is rumoured to be heading for a stock market listing.

Most interesting is to look at both the ambitions and direction of Tesla and xAI specifically. The two companies under Musk’s control appear to be overlapping more and more. In materials obtained by Bloomberg, xAI told its investors that the goal is to build an autonomous AI system that could be used to control humanoid robots — like Tesla’s Optimus.

Tesla has received a great deal of attention for these robots, but it should be noted that they are one of many players in this market. China’s Unitree Robotics and South Korea’s Hyundai are two major competitors. The latter are also working with Google’s AI lab DeepMind to use their AI models to control the robots.

One might also wonder how Tesla’s ambitions in this area are actually progressing. In May last year, Musk said he expected to produce thousands of Optimus robots, and that they would be working in Tesla factories before the year was out. That plan was subsequently abandoned last summer.

xAI’s strategy raises some questions about how far Tesla has actually come when it comes to a functional, smart, AI-driven Optimus robot. To Tesla’s shareholders, Musk has said he believes the company can become the world’s most valuable company by a wide margin. They will not get there with electric cars alone.

Musk has therefore started talking more and more about self-driving cars and robots as new areas for Tesla to grow into. But if xAI — an entirely separate company — intends to be the brain behind Optimus, what does Tesla contribute beyond manufacturing the robots?

Tesla — like Apple with the iPhone — tends to talk about the advantages of what is called “vertical integration.” In simplified terms, this means that as a company you own many steps in a manufacturing process.

For both Tesla and Apple, this means producing both hardware and software, for example. Through that, they can create a type of experience that is difficult for others to compete with. For Tesla to outsource the robots’ “brain” to another company would therefore be unusual for them.

But there is a conceivable solution.

Everything therefore points towards a consolidation of Musk’s corporate empire. If Tesla wants to step back somewhat from ordinary electric vehicles in terms of positioning, the most logical move would be to acquire xAI. Ideally Musk would probably also like to bring SpaceX into this new corporate structure, but that would likely be too complicated.

At Tesla’s recent annual meeting, many shareholders expressed support for an investment in xAI. But why not merge them entirely? It would move Tesla faster towards robotics, give it a stronger position in the red-hot AI market — and retain Tesla’s strongest card: its stock market position with engaged shareholders who look past the enormous valuation.

Such a merger would bring robot manufacturing and AI development under the same roof. Elon Musk could then return to doing what he does better than anyone: selling the future to the stock market.

He has probably already forgotten the bikini images. They were just a minor blip on the curve for the new supercompany Musk wants to build.

Small investors were going to get rich — what happened?

December 22, 2025
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on December 22nd, 2025. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

Over five years, more than 40 percent of the value in Tin Fonder Ny Teknik has disappeared. In the past year alone, over 2.5 billion kronor has flowed out of the heavily criticised fund. It was supposed to ride the Swedish tech miracle — and make ordinary savers rich. What went wrong?

“We want to take on that leadership jersey, help these companies to the stock market and make sure the companies stay in the Nordics.”

Carl Armfelt, one of two fund managers at newly launched Tin Fonder, gives an interview to Dagens industri in February 2019.

Together with Eric Sprinchorn, he is about to launch the company’s first fund. The duo has broken away from Swedbank Robur and its fund Ny Teknik. Now they want to do something similar: invest in smaller Nordic tech companies.

“We believe this type of company is the future,” says Sprinchorn.

Nearly seven years later, a partial accounting is in order. How did that future Swedish and Nordic tech miracle actually work out? And how well does the leadership jersey fit Tin Fonder?

The phrase “the Swedish tech miracle” had spread widely the previous year — 2018 — when journalist Kina Zeidler published a book with the same title. She interviews investors, researchers, and employees at tech companies to tell the story of “how little Sweden became the best in the world at producing global successes.”

As early as 2013, The Economist is on a similar trail, and confidence begins to grow in Sweden. For a relatively small population of Swedes, a disproportionate number of Swedish tech companies are doing extraordinarily well. There seems to be something going on.

In this environment, it is easy to understand why new funds are launched to capture this phenomenon. The successes have grown so large that even ordinary savers have taken notice. Everyone listens to music via Spotify and buys on Klarna invoices, while the kids build digital worlds in Minecraft. But not all Swedish tech companies share this positive trajectory. Far from it.

And that turns out to be a problem when you want to be a fund that takes small savers along for the ride.

To capture value growth, it helps to come in early as an owner. The company needs to be mature enough for an investor to understand what it wants to do, but not so late that investing becomes too expensive. Carl Armfelt’s comment at the start about helping companies to the stock market speaks to this. With Tin Fonder as a partner, Swedish and Nordic tech companies should be able to list on the stock exchange.

One way to be such a partner is to become an “anchor investor” — guaranteeing a certain proportion of investment and ownership before a company begins trading on the exchange. Tin Fonder does this on several occasions over the years.

In 2019, for example, they invest 10 million kronor in gaming company Moba Network, taking 39 percent of the total share issue. The company has a shaky start, but the investment is defended by Armfelt in an interview in DI shortly afterwards. He compares it to another gaming company Tin Fonder had invested in — Stillfront. In the interview, Armfelt notes that Stillfront had listed five years earlier and risen tenfold since then.

With hindsight, however, both investments are something of a minor disaster.

Moba Network has lost over 93 percent of its value since listing. Stillfront — after a gigantic peak around 2021 — has fallen 10 percent since 2016 and added a massive debt pile to its balance sheet along the way. The share price now stands at six kronor, compared with 119 kronor at its peak.

In 2021 the strategy continues and Tin Fonder Ny Teknik goes in as anchor investor in yet more companies.

The figures since each listing speak for themselves. Swedish survey company Cint Group: minus 97 percent. Identity company Checkin.com: minus 86 percent.

Even if Tin Fonder has sold off some of these failures, there are examples of dismal performance among the holdings that remain in the portfolio. Fractal Gaming listed in February 2021 alongside Tin Fonder and has since almost halved in value.

Looking at more recent times, it is not much more cheerful.

The fund’s largest holding, Surgical Science, has lost over 78 percent of its value this year. Another, Danish Novo Nordisk, has shed half its value. The Danish pharmaceutical giant is now trading at a similar price to before Ozempic and the other weight-loss products were launched.

Tin Fonder has received a great deal of criticism for delivering poor returns while charging high fees. But is it just a matter of poor management and backing the wrong shares — or are there other factors that explain the weak performance?

If you look at Swedish tech funds more broadly, you quickly get an entirely different picture.

Armfelt and Sprinchorn’s former employer, Swedbank Robur, runs the popular fund Technology. Over the past five years, Tin Fonder Ny Teknik has lost around 42 percent of its value, while Robur Technology has increased by 149 percent.

How can two actively managed tech funds based in Sweden have such dramatically different outcomes?

The answer is quickly found by looking at Robur Technology’s larger holdings. The largest is Microsoft, then Nvidia, Broadcom, Apple, and Taiwanese chipmaker TSMC. The fund is 90 percent US-based. The equivalent figure for Tin Fonder Ny Teknik is 4.6 percent in the US. The American tech giants have left behind not just Swedish tech stocks — but the entire stock market, across all categories.

A fairer comparison is therefore to look at Robur’s other tech fund, Ny Teknik — the one the Tin founders worked with before. Over five years that fund has also fallen by over 35 percent, while the Stockholm stock exchange, OMX30, has risen 45 percent over the same period.

The Swedish tech miracle has therefore not delivered particularly well on the Swedish stock exchange. And ordinary savers are starting to notice. The outflow from Tin Fonder Ny Teknik over the past twelve months stood at over 2.5 billion kronor. The strategy of betting on smaller Nordic tech companies is not working right now, at least.

At the same time as Tin Fonder launched its business, the foundations were laid for a broader misunderstanding about Swedish tech — one that persists to this day, in LinkedIn cheering sections and in Sweden’s own digitalisation strategy alike.

Namely, that Sweden should be some kind of unique unicorn factory that has created more billion-dollar-valued companies than any country except the US and Silicon Valley. That is not true today, but there may have been a brief period when we as a country held a podium position.

A more accurate picture of Swedish tech would be to say that things have gone fantastically well for quite a small number of companies. A similar pattern can also be seen globally, just on a considerably larger scale.

Individual companies achieve extraordinary success, while a great many others limp along.

But the picture is not entirely bleak for Swedish tech companies on the stock exchange. Someone who invested in Spotify five years ago has seen a 77 percent increase. Gaming company Betsson has risen a full 95 percent. Neither is in the same league as global shooting star Nvidia — up 1,330 percent — but the performance is strong.

So yes, Swedish tech can do well on the stock market. But the success is very unevenly distributed. And savers in funds that have bet on this stock market category have not seen much of it either.

Despite the dismal performance on the stock market for many Swedish tech companies, it is possible to find winners in the segment — even outside the equity market. In the pay packets of some of the fund managers, for example. During the period 2020–2024, the owners of Tin Fonder drew over 240 million kronor in share dividends.

That, if anything, is a tech miracle.

A Big Short moment for AI?

December 20, 2025
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on December 20th, 2025. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

Hundreds of billions have been made by those who bet on Oracle as a winner in the AI race. Now the opposite is growing — investors who see a chance for profit if the tech company can no longer service its loans. Some are warning of a possible bomb.

A long table in dark wood runs through the conference room. On one side sits Michael Burry, played by Christian Bale, and on the other, representatives from investment bank Goldman Sachs.

The scene from the film The Big Short has become something of an icon. Burry convinces the bank to sell a financial product that pays out if a large number of mortgages cannot be repaid. The banker explains that what Burry is asking for would be seen as a very bad investment by most in the outside world.

Burry replies quickly — and it later turns out to be entirely correct: “Everyone else is wrong.”

The product Burry buys is called a CDS — a “credit default swap.” You can think of it as a form of insurance product, in that it pays out when a loan cannot be repaid.

But unlike insurance, you do not need to own the underlying asset to buy a CDS. You can profit from other people’s misfortune regardless.

And that is precisely what has started happening in the bubbling world of AI.

Larry Ellison has become increasingly prominent and well-known in recent years. As the founder of the previously rather sleepy database company Oracle, he had long been a billionaire — with luxury yachts and the classic trappings that come with that life. The company was stable, profitable, and a little uninteresting to most. Until around 2021, that is. That is where the next great upward journey began — and into the inner circle of the tech world.

Ellison’s Oracle is a partner to TikTok on data storage, and is set to become a part-owner of its American operations. But it is with the AI wave that Oracle has truly come into its own. All these data centres needed to expand capacity have to be built by someone — and Oracle has stepped forward. Data centres are something they have built before given their cloud business, but the scale is now something else entirely.

This week it emerged that Oracle’s lease payments for data centres yet to be built amount to 248 billion dollars — around 2,300 billion kronor. These are to be paid between now and 2028. The figure Oracle reported the previous quarter was around 100 billion dollars. The increase is somewhat astronomical, and analysis firm CreditSights described the situation as a “bomb.”

The question the market is now starting to ask is what happens if Oracle cannot make these payments. And shortly afterwards, people start wondering whether there is money to be made if Oracle fails to deliver. That is where the CDSs come in.

Since September, trading volumes in these products have increased by 90 percent. For CDSs linked specifically to Oracle, volumes have tripled during the year and are now trading at their highest price in over 15 years.

This does not necessarily mean the market believes Oracle will go bankrupt. But it clearly shows that many are concerned the company may have taken on too many large, expensive projects at once — and that investors want some form of protection or insurance should things start to go wrong.

Many AI companies are building data centres at the moment, so the situation is not unique. Microsoft, Amazon, and Meta all have major projects under way in the area. The problem here is the scale.

Oracle’s commitments are larger than all three of the other tech giants’ combined. The company itself estimates that nearly 75 percent of Oracle’s revenues in the coming months will be spent on data centre expansion.

The concern also sits one level above Oracle — with its customers. OpenAI has made a deal worth 300 billion dollars with them, for example. The money is to go towards data capacity over five years, starting in 2027. For Oracle to be able to deliver on this, they need to build.

But there is another problem: at present, OpenAI does not have all 300 billion dollars to pay with. That must be resolved before the payments are due.

Given the situation, it is easy to understand why there are people thinking like the investor Michael Burry from The Big Short. Do all these AI investments actually add up? And if they do not — how can I make money from the fact that everyone around me is wrong?

The last time the price of a credit default swap for Oracle was this high was 2009.

And most people probably remember what happened in the financial world back then.

The most powerful player is never held to account

December 16, 2025
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on December 16th, 2025. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

The SVT documentary “Hatet” about Joakim Lundell misses the most powerful player in its investigation. The actor that enables threats and the spreading of rumours is not asked a single question.

In ancient Rome, gladiatorial games were a popular form of entertainment. In the arena, fighters battled both each other and wild animals to see who would emerge victorious — and alive — from the spectacle.

It is hard not to think of gladiatorial games when watching the third and final part of SVT’s documentary “Hatet.”

Here an obvious human tragedy unfolds before a public audience, but SVT appears to have missed the arena in which it takes place. Back then it was the Colosseum, with an emperor dictating the terms. Today it is YouTube. Why are no questions asked of those who have the most to gain from this brutal entertainment?

We have come a long way from the entertainment monopoly that SVT once held. Not so long ago there were two TV channels to choose from — both from SVT — and you watched what was broadcast. Today the competition is something else entirely. You watch when you want, content ranging from 10 seconds to three hours — from TikTok to SVT Play to YouTube and back again.

Some of this media logic appears to have stuck at SVT. Why else would you make three hours of documentary about an influencer family and its feuds? Because they are famous. And as former SVT journalist Anna Hedenmo notes in Expressen: they want to reach a young audience. “If you can’t beat them, join them,” as the saying goes.

The irony is therefore striking when you have worked through these three hours and realise that at no point is a single question put to the platform where all of this takes place.

The conflict between brothers Joakim Lundell and Christofer “Chrippa” Lundström has been called the biggest internet drama in Swedish history. So big that the subject has also become the focus of countless videos from other creators. YouTube is mentioned constantly — it is the family’s primary source of income and channel of communication — but the company is not asked any questions.

A large share of all the threats and accusations shown in the documentary take place on YouTube, where the format is very simple: controversies work. More clicks lead to more advertising revenue. There is a reason why video titles with nuance are not popular. The headlines that attract viewers are screaming, in capitals, with garish images.

YouTube is not alone in this commercial logic. But unlike other journalistic products, there is no responsible publisher, and the platform does not explicitly approve clips before they are posted.

The advertising revenue does not only benefit Joakim Lundell, Christofer “Chrippa” Lundström, Pontus “Anjo” Björlund, and the others in the documentary. The single biggest winner is YouTube — economically too.

Without lifting a finger, the family drama drives millions of ad impressions and cements the service as an entertainment hub to visit daily. The incentives not to intervene in the content, regardless of what consequences it has for those affected, are therefore very clear. YouTube’s automated systems are designed to scan everything that is published and flag up inappropriate material. But the system is far from perfect, and many videos slip through anyway.

In the documentary podcast Badfluence from SvD and Podme, 16 examples of video clips containing potential violations were sent to YouTube. After review, one of them was taken down and another appears to have been edited. But this only happens, it seems, when they are specifically brought to attention by the media.

The standard explanation from YouTube is usually that so much material is uploaded that it is not possible to review everything manually — and that a great many clips are never allowed to be published in the first place. But anyone who has spent more than ten minutes on YouTube knows that these grey zones are extraordinarily large. And therefore also very profitable.

An obvious thing to do, therefore, would be to hold the arena where all of this takes place to account. Why does SVT not ask YouTube any questions? Why do they not see that it is the site’s algorithms that encourage the behaviour shown in the documentary?

When SVT surveyed this new media landscape, they identified Joakim Lundell as a new power player — popular with the coveted young audience. So far so correct. But they appear not to understand the context in which he operates — how one of the world’s largest tech companies provides, encourages, and indirectly pays for these controversial videos on its platform.

In their eagerness to accommodate young viewers, they forgot to scrutinise — or even ask questions of — the most powerful player in the industry.

“Hatet” makes YouTube appear to be neutral ground. It is the absolute opposite. Almost everyone knows this. Had SVT asked any of the young people they are trying to reach, they could have explained how it works.

Netflix’s masterstroke — a win even in defeat

December 9, 2025
by Björn Jeffery
SvD Näringsliv

This analysis was first published in SvD Näringsliv, in Swedish, on December 9th, 2025. This piece was translated from Swedish by Claude. Some phrasing may differ from a human translation.

What would you pay to hobble your worst competitor? Netflix’s answer to that question is 54 billion kronor. And should Warner Bros pull out of the deal — Netflix gets a sizeable Christmas bonus either way. The losers, whichever way it goes, are viewers.

The American media world currently resembles a middle-school disco.

First nobody wants to dance. Then everybody wants to dance — but with the same person. You might have expected deals in the hundreds-of-billions bracket to be a little more sophisticated than that, but no.

Warner Bros CEO David Zaslav has agreed to sell HBO and Warner Studios to Netflix. Immediately afterwards came a higher, hostile bid from Paramount Skydance. It is going to get messy. But everything points to Netflix being the big winner — whatever the outcome.

To understand Netflix’s strategy behind the mega-deal, all you need to do is look at one particular point in the purchase agreement. It is called a “breakup fee” — a sum of money that the buyer, Netflix, pays to the seller if the deal does not go through.

For large deals of this calibre — what Warner Bros is selling is valued at over 780 billion kronor including debt — there is significant political risk. Competition issues are sensitive, particularly when it comes to media.

For TV companies the matter would have been handled by the FCC — the Federal Communications Commission — but since this deal does not cover ordinary TV channels, it falls outside their remit. Instead it will be the US Department of Justice and the FTC, the Federal Trade Commission, that will scrutinise the deal.

And one more person, of course. Donald Trump.

His son-in-law Jared Kushner is investing alongside Paramount Skydance, which in turn is owned by Trump’s friend and supporter Larry Ellison. Warner Bros also owns the TV channel CNN, which is not Trump’s favourite. There are many relationships that could come to play a role in this decision.

If any of these political institutions were to block the deal, Netflix would need to pay a breakup fee of around 54 billion kronor to Warner Bros. That might sound like a lot, but one should bear in mind that Netflix has a market capitalisation of just under 4,000 billion kronor. And before you judge whether something is expensive, you should look at what they get for the money.

Political processes of this nature are long and drawn-out. If we look at another major media deal involving the same company — when AT&T bought Time Warner — it took 20 months before it was concluded. A reasonable guess is that this could take at least a year, if the deal is not called off beforehand.

During that time, HBO and Warner Studios will be in limbo. No major investments and changes are likely to happen while they await the new owners. A great deal of management’s time will be spent trying to get the deal approved rather than developing the business. For a market leader like Netflix, that is worth its weight in gold. Or rather — it could well be worth 54 billion kronor.

If the deal goes through and Netflix gets what it wants, they will have become the overwhelmingly dominant player in the streaming market. If the deal does not go through, they will have paid a little over one percent of their market capitalisation to ensure that their main competitor HBO has wasted a year on political paperwork.

But there is yet another way this could end. Should Warner Bros decide to accept Paramount Skydance’s new, higher bid after all — they would also need to pay a breakup fee, but in the other direction.

Netflix would then receive half — 27 billion kronor — as a consolation prize for Warner changing its mind. A decent Christmas bonus? The whole thing is very elegantly played by Netflix’s management, with CEO Ted Sarandos at the helm.

For Netflix or HBO viewers, nothing will be noticeable for the coming year at least. These are two separate services that will try to sign you up as a subscriber, just as before. But looking at a slightly longer horizon, the number of streaming services will in all likelihood decrease. Paramount Skydance clearly has big ambitions and will not give up even if they fail to push through this particular deal. Fewer services means less competition, which in turn could lead to higher prices. Consolidation of this kind rarely benefits viewers.

The winner in all of this — somewhat regardless of how it plays out — therefore looks set to be Netflix. But before they celebrate too much, one should carry a little media history into the room. Enormous deals in this category tend not to age particularly well. The aforementioned AT&T and Time Warner is one example. AOL and Time Warner another. An extraordinary number of billions have gone up in smoke when excessive confidence in the future has been allowed to steer.

And it seems Netflix is not immune to that either.