Media War: welcome to the dark side

There is a shelf in my very much shelved office where books slowly stop being books and start becoming warning systems. You know the kind: grim, dark, and not messing around. The ones you keep returning to because the world insists on turning them from analysis into documentary footage. 1984 and Animal Farm stand here, Orwell wrote them as a warning, not a manual. Chris Miller’s Chip War lives there. Amy Webb’s convergence madness lives there. A few dusty cyberpunk prophets sit at the end of the row, looking increasingly annoyed that nobody listened to them in time. Bruce Sterling, Asimov, William Gibson. The shelf is east-facing, which feels appropriate. It’s gunmetal steel. The sun hits these books first every morning, and they all radiate the same low bad-humored hum: you were warned. Yep, my office is a threat.

Yesterday I added a new spine to it (which is the holy moment a book navigates from the nightstand to the office after multiple reads). Right next to Chip War, Martin Andree’s Media War. (The English edition dropped the s. My shelf keeps it, because in my head the two books are a matched pair and pairs need symmetry. So does my brain.) That placement next to that other black book matters more than it looks.

Chip War explained how semiconductors became geopolitics. Memory, fabrication capacity, lithography, all of it quietly turning into strategic weapons. The book reads like a thriller until you realize half the planet’s future runs through a Taiwanese island, a handful of Dutch machines, and whether Washington wakes up grumpy on a given Tuesday because the reflecting pool has the wrong shade of billionaire blue. Media War does something equally uncomfortable. It explains how digital media infrastructure stopped being infrastructure and quietly became power itself. Past influence, past reach, all the way through to actual power.

The interview I did not expect to be this dark

Yesterday I sat down with Andree in the Solvay Library in Brussels for a podcast interview before the BPX conference. Gauthier Elslander, MD of BPX (the Belgian Publishing Experience, the joint platform of DPG Media, Mediahuis, Roularta, Rossel and IPM), chose the location of his conference extremely well. If you have never been there, picture a room built in 1902 for the kind of nineteenth-century industrialists who genuinely believed they were arranging the future. Wood, glass, books up to the ceiling, the soft light that European institutions reserve for moments they want photographed. We sat in a building full of leather-bound volumes that probably outdate Belgium itself, in a room dedicated to Einstein, and proceeded to have a forty-minute conversation about how democratic legitimacy is being quietly mortgaged to a dozen American corporations.

Cheerful stuff over a chilled drink. I had read Media War with what I would politely call enormous attention (nudging closer to 60 than ever before in my life, I continue to have a sacred fondness of reading the real books rather than blazing through the ChatGPT three-paragraph summary). By page forty I had filled two mental notebooks. By page one hundred I had stopped sleeping properly. The reason I look so forward to meeting Andree in person was not to get the soundbites (he has plenty of those). It was to find out whether the man behind the moody manuscript was as uncompromising in person as he is on the page. He is. If anything, the live version is sharper, and dangerously quick. He talks the way he writes: no diplomatic upholstery. No conference-circuit hedging. No “on the other hand” softeners designed to keep the panel sponsors comfortable.

Frankly, thank God. We already have enough carefully massaged LinkedIn sludge written by corporate diplomats (let me rephrase that: by their bots), pretending everything is complex and nuanced while Silicon Valley reverse-engineers feudalism with better UX, free shipping, and a multi-colored ribbon smelling proudly of subtle Baccarat Rouge 540.

The sentence that should make every European policymaker choke on their croissant

Early in our conversation Andree dropped a line that has been bouncing around my head since. He was explaining the structural problem with the platform economy, and out came this:

“Under conditions of monopolies, freedom turns into its opposite.”

There it was, the whole book in one sentence. The whole decade in one sentence, if I am being honest. For twenty years we have been sold a fairy tale. The internet as liberation machine: open, democratic, horizontal, diverse, communities, participation. On top of that: connection, sharing, smiling non-acneed millennials on stock photos holding expensive laptops in soft-lighted cafés or sunny beaches for reasons nobody can fully explain. Meanwhile, a microscopic handful of corporations quietly absorbed the digital nervous system of humanity. Andree’s own Cologne baseline study puts the numbers on the table, and they are not very subtle: Google around 90% of search, YouTube around 80% of free video, Meta around 85% of social media. That is the infrastructure layer of opinion formation in a democratic society, owned by three companies headquartered between Mountain View and Menlo Park.

That is no longer a market. That is terrain occupation. So far the illusion of “free speech”.

They were not corrupted. They were always like this.

The thing I appreciated most about the conversation was that Andree refuses the comforting bedtime story that Big Tech somehow went bad. He dismantles that fairy tale almost with irritation. They were never benevolent hippies who drifted toward authoritarian monopoly power after eating too much venture capital, ketamine, and reading too much Ayn Rand at Burning Man. The anti-state libertarian DNA was always there. The mythology simply worked astonishingly well. His version of the origin story is worth quoting at length:

“The early tech corporations always told a fairy tale. It goes like this. The internet is cool and free. It’s anarchic. Come in with your communities and your participation. And the state is evil. The state is horrible. They want to control the free internet. They want to take it away from us. If you listen very carefully, it’s a story that denies the sovereignty of the democratic state in the digital world.”

He is not arguing that the platforms slid into anti-democratic positions over time. He is arguing that anti-democratic positioning was the founding theology of the entire sector and we politely chose not to see it. He name-checks David Golumbia, who was making essentially the same argument as early as 2010 and got the standard treatment European academics get when they are right too early: ignored, then forgotten, then quietly cited fifteen years later when the ceiling falls in.

I have been in tech long enough to remember the early speeches. The TED talks, the conference keynotes. The breathless coverage, the genuine, sincere belief from a lot of very smart people that this time the new thing was going to be different and better and flatter. And Andree’s diagnosis sits there like a hand grenade without a safety pin in the corner of the room: we collectively mistook convenience for freedom. That little linguistic confusion may end up costing us something a great deal more expensive than a Prime subscription.

Dark Tech, Dark Enlightenment, and why Andree picked his words on purpose

I asked him about the very loaded language he uses in his book. Weaponizing media. Dark Tech. Dark Enlightenment. It sounds, when you say it out loud in a library in Brussels, like the kind of vocabulary that gets you politely uninvited from the next panel. He pushed back, hard:

“The vocabulary of war is not my invention. It is the lexicon used by the actors that are currently doing this. The word ‘dark’ is not a creation from my side. The ideological scripts that this ecosystem is using are openly invoking the notion of Dark Enlightenment.”

The reference is to Nick Land’s Dark Enlightenment, the foundational text of the neo-reactionary movement that proposes, with a straight face, that democracy is an obsolete operating system and should be replaced by corporate sovereigns running city-states. It is not a marginal document. It is the philosophical scaffolding behind a significant chunk of the Peter Thiel universe and a lot of the people who currently have meetings at the White House. Andree’s point is that the language is not metaphor. The actors are using these words about themselves. He is just refusing to soften them for European audiences who would prefer the conversation stay civil.

The PayPal Mafia, where the philosophy met the checkbook

If you want to understand how a fringe accelerationist essay turned into actual American economic and political reality, you have to read Jimmy Soni’s The Founders: The Story of PayPal and the Entrepreneurs Who Shaped Silicon Valley. Soni was given full access to the cohort, and the book sits on my shelf one row below Media War and Chip War for a very specific reason. It is the missing middle chapter. Nick Land wrote the philosophy. The PayPal Mafia turned the philosophy into a capital allocation strategy. That capital allocation strategy is now writing American AI policy, American foreign policy, and increasingly, American constitutional interpretation.

The cast list, for those who have managed to avoid it: Peter Thiel (the “Don of the PayPal Mafia,” founder of Palantir and Founders Fund, first outside investor in Facebook, godfather of Vice President JD Vance). Elon Musk (X.com, then Tesla, SpaceX, X, xAI, and a brief detour as cost-cutter-in-chief of the United States federal government). David Sacks (founding COO of PayPal, now AI and crypto czar at the White House). Reid Hoffman (LinkedIn, Greylock, the only one of the cohort to actively fund the Democratic side of the same coin). Keith Rabois. Roelof Botha. Max Levchin. Luke Nosek. Ken Howery (Thiel’s old roommate, now US Ambassador to Denmark, which is doing wonders for Greenland diplomacy). And further out in the orbit, the YouTube founders (Chad Hurley, Steve Chen, Jawed Karim), Jeremy Stoppelman of Yelp, and most of the early Yammer crew.

Twenty or so people in a 2007 Fortune cover photo dressed in mafioso costumes, half-joking about the nickname. They have since founded, funded, or served as executives for YouTube, Facebook, SpaceX, Tesla, Lyft, Airbnb, Palantir, Pinterest, OpenAI, Reddit, Eventbrite, Tumblr, LinkedIn, Quora, Yelp, Postmates, Square, and Stripe. Roughly speaking, the consumer internet. Roughly speaking, the AI race. Roughly speaking, the venture capital flowing into roughly all of it. The 2007 photograph aged the way a wedding photograph ages when one of the spouses later turns out to be a contract killer: nobody is laughing anymore.

The Soni book is fascinating because it lets you watch the worldview congeal in real time. Thiel’s The Diversity Myth from 1995, his Stanford circle, his Zero to One thesis that “competition is for losers” and that monopoly is the natural and desirable end state of any serious business. Read that line again with Andree’s argument in the background. Competition is for losers. Monopoly is the goal. It is, almost word for word, the inverted version of every antitrust principle Europe has been trying to operate on since the 1950s. And the man who wrote it now has a closer relationship with the sitting administration than most Cabinet secretaries.

He is not, of course, the only one. He is just the most legible one. The point Andree keeps making, and which Soni’s book accidentally illustrates with the warmth of an authorized biography, is that the Dark Enlightenment is not a fringe internet movement. It is a staffing plan. And the people doing the staffing are the same names you would have found on a Y-Combinator demo day fifteen years ago.

He also kept returning to a quieter idea, which is the one I find genuinely frightening. The Dark Tech project, in his reading, is not primarily about money. It is about replacing democratic legitimacy with technological legitimacy. The engineer as sovereign. The founder as philosopher king. The platform as state. The algorithm as invisible governor. The board meeting as constitutional convention. Sounds excessive? Give it five years. I have a calendar reminder for May 2031 that just says “check on Andree’s predictions”. I do not expect to enjoy reading it.

The trap (and why “just leave the platforms” is a fantasy)

Halfway through our conversation Andree handed me the cleanest articulation of the platform trap I have ever heard. He was responding to the standard accusation that he should not sell his books on Amazon if he hates Amazon, which is the kind of bad-faith point that gets tossed at every serious critic of Big Tech the moment they go on a press tour. His answer:

“89% of digital book sales happen on Amazon. So I would lose all this audience if I wouldn’t use Amazon at all. That also shows you how deep we are in the mess.”

There is the trap. Under monopoly conditions, opting out becomes performative theater. You can leave X. Wonderful, your audience stays there, your market stays there. Your competitors stay there. Your cultural relevance stays there. The infrastructure stays there. You remain orbiting the monopoly whether you like it or not. Andree’s own move, which I find very honest (and which is also my very personal strategy), is to use both. He publishes on the alternatives and on the monopoly platforms, because shouting into a 0.8% reach silo is exactly the outcome the platforms want. (As he put it, “they will laugh like hell because you stay in your little whatever thing”. The man does not mince.)

This is the part I want you to sit with for a moment. We talk about digital sovereignty in our boardrooms and on our stages the way medieval lords talked about chivalry: with great feeling and very little practical effect. We run cloud independence reviews. We commission risk registers. We add strategic supplier diversification to our slide decks. We preach free media choice. And then we sign another three-year enterprise contract with the same three vendors because there is genuinely nowhere else to go. The lock-in is not contractual. The lock-in is structural. The platforms ate the ground we were planning to stand on.

And then the AI layer lands on top

The truly eerie part of the conversation was when Andree pivoted to AI. I asked, half-hoping for reassurance, whether the AI race might at least reshuffle the deck. His answer was unsentimental:

“The AI market will affect the gigantic market of services. That is why the US is investing hundreds of billions of dollars. The one who catches it again will win the race, and then you will have again gigantic monopolies that cannot be attacked anymore in competitive terms.”

So no, nope, nein: the deck is not being reshuffled. The same hands are dealing themselves a second deck. Search monopolies were already dangerous. Social monopolies already destabilized democracies (and arguably swung at least two national elections in the last decade). AI becomes the meta-layer on top of all of it. Whoever controls foundational AI infrastructure controls discovery, mediation, productivity, advertising, recommendation, information flow, enterprise tooling, automation, and, increasingly, decision-making itself. The AI race is not a tech race. It is a planetary land grab with GPUs and a power agenda.

Europe, as ever, is having a deeply sophisticated discussion while the Americans and the Chinese are already pouring concrete. Andree was particularly sharp on the EU’s hesitation. The regulatory tools exist. The Digital Markets Act exists. The Digital Services Act exists. What is missing is political nerve. Berlin hesitates because Berlin is terrified of Trump-era tariffs. Brussels hesitates because Berlin hesitates. The fines, when they happen, are an annual rounding error in the platforms’ tax provisions. Meanwhile the platforms continue to integrate themselves more deeply into every layer of European life, including, with delicious irony, the layer that is supposed to be regulating them. O tempora, o mores…

Classic Europe. Invent the regulation. Lose the momentum. Hold a summit. Produce a PDF. Schedule another summit. Somewhere in Palo Alto, a tech executive buys another democracy for parts.

The personal bit, because this is a personal blog

I left the Solvay Library yesterday a little quieter than I went in. Forty minutes with Andree did not so much teach me anything new. It did something worse. It rearranged things I already knew into a pattern I had been carefully avoiding. I have spent thirty years en grand écart between the tech and the influence industries. I have built media strategies on top of these platforms. I have advised clients to partner with them, to embed with them, to prioritize them, to measure themselves by their APIs. I have written posts using the very algorithms Andree is describing as the operating system of authoritarian drift. I am, by any honest accounting, complicit. And I do not plan to stop.

The complicity is uncomfortable, but that is a comforting feeling. It is also, I think, the only useful starting point. Pretending I have been a clean observer of this from the sidelines would be both dishonest and useless. Andree’s argument lands harder when you accept that we built this. And by we I do not mean the engineers in Mountain View. I mean us. The media people, the consultants, the CIOs, the agency heads, the conference speakers, the LinkedIn thought leaders, every executive who ever stood on a stage and said digital transformation while quietly handing the audience’s attention to a couple of deep tech companies. Let’s use them for good, for better, for best. Very conscious of what they are, very conscious of what they can do. Very, let me plant a concept of the great Belgian philosopher Jean-Claude Van Damme, aware.

Which brings me to the conclusion Andree refuses to soften, and which I will not soften either. Big Tech Must Go (the 2023 book that won him the Günter Wallraff Special Award for Press Freedom and Human Rights) was the warning. Media War is the post-mortem on a patient who is still technically breathing. The fix is well-understood, almost embarrassingly so: open links, open standards, infrastructure-content separation, market-share caps, content liability, more chances for alternatives and local players. He has been listing these for years. They are not radical, they are boringly orthodox antitrust. The reason they have not been implemented has nothing to do with technical difficulty and everything to do with political fear and practical reach.

The shelf, again

Walking back to my car last night I was thinking about that east-facing shelf in my office. Chip War and Media War now sit side by side. One book explains how compute became power. The other explains how communication became sovereignty. Together they form a rather nasty little duet about the same uncomfortable century-defining fact: the infrastructure layer of modern life is owned by a handful of companies that do not particularly believe in democracy, and Europe is still trying to decide whether regulation is rude. They fit well with their neighbor The Coming Wave from Mustafa Suleyman.

There is a fourth book that probably belongs on the shelf next to those three. I have not written it yet. I am not sure I am the right person to. But if anyone is going to, it will be someone like Andree, and the title will probably need an even darker word than war.

In the meantime, buy his book. Yes, on Amazon if you have to. He already made peace with the irony. So can you. Very civilized of us, very elegant.

Very potentially catastrophic.