Gil Durán has spent the last few years doing something useful: listening when Silicon Valley’s strangest political figures tell us what they want.
Increasingly, they want government.
His forthcoming book, The Nerd Reich: Silicon Valley Fascism and the War on Democracy, follows an anti-democratic current running through people and projects associated with Peter Thiel, Marc Andreessen, Balaji Srinivasan, Elon Musk, Curtis Yarvin, the Network State, and the older dream of technological exit from ordinary democratic constraint. Durán’s argument is deliberately severe: some of the richest people produced by the current order have moved beyond complaining about government toward imagining political orders in which people like themselves govern instead.
I am not going to spend much energy deciding whether the CEO monarchy is a good idea.
It is not. We can save some time.
¢¢¢¢
There is something almost generous about an explicit authoritarian.
That guy tells you where to look.
If someone says “democracy is obsolete”, that “a country should be run more like a corporation”, that “exit should replace public obligation”, or that “superior people should have greater political authority because they are superior”, the constitutional problem arrives wearing a name tag.
Durán’s subjects have provided plenty of material in this category. Yarvin has long argued against democratic government. Srinivasan’s Network State imagines digitally organized communities acquiring increasing political and territorial reality outside familiar state structures. Durán connects these ideas to a broader Silicon Valley politics in which technological and financial capacity becomes evidence that existing democratic institutions are impediments to be escaped, captured, or replaced.
Fine. Political theory knows this guy already. The king has better branding now. He owns GPUs. His court has a podcast. We can work with this.
The Nerd Reich is the easy case because the sovereignty claim is very visible. Somebody wants authority. Somebody dislikes the institution that currently limits it. Somebody has a theory about why the people holding extraordinary capacity should receive extraordinary jurisdiction.
They have done us the courtesy of putting the crown on the table.
Now, try taking the crown away.
¢¢¢¢
Excellent. What did we just solve?
Artificial intelligence is still extraordinarily useful.
Now give me the nicest artificial-intelligence company in America.
Better yet, let the system actually work.
I want this. That is why this is the hard case.
Every success earns the next permission for a perfectly defensible reason.
People learn the interface. Old systems are retired. Records accumulate in its format. Budgets move toward the service. New employees are trained around it. Entire workflows begin to assume that it will still exist tomorrow.
Eventually, replacement stops looking like ordinary competition.
It looks dangerous.
The operator does not need to say, “I should rule you.”
It can say:
And the worst part is that these statements may be true. A lie is not even required.
¢¢¢¢
That is the larger problem hiding behind the Nerd Reich.
Sovereignty does not require a sovereign personality.
An institution can begin with a function everyone has good reason to preserve. As it becomes better at carrying the function, reliance grows. As reliance grows, replacement becomes harder. Once replacement becomes sufficiently dangerous, the institution acquires a new kind of power: it can use the importance of the function it carries as an argument for the continuation of its own control.
Nobody has to repeal the constitution.
Nobody has to suspend elections.
Nobody has to declare himself king.
The legislature continues meeting. Courts continue issuing opinions. Agencies continue operating. Elections continue producing winners and losers.
Meanwhile, the practical routes through which those institutions know, coordinate, remember, authorize, and act can converge on one layer that none of them can easily replace.
At that point, changing the people at the top does not settle the question.
Neither does nationalizing the system.
Public ownership may improve accountability, distribute gains, secure access, and remove some forms of private capture. Those are real advantages.
A state can still build an arrangement in which identity, evidence, coordination, infrastructure, and review all travel through institutions whose correction ultimately returns to the same center.
The question is larger than who sits in the chair.
It is whether one chair even has to exist.
¢¢¢¢
Durán’s subjects are useful again here because they say loudly what a much larger institutional process can accomplish without saying anything at all.
The temptation is simple:
The authoritarian has an answer.
The founder has a softer version.
The military has one.
The expert has one.
The platform has one.
Each argument can identify a real function.
Each can also promote contact with that function into title over the conditions under which everyone else may correct it.
That promotion is where the political problem lives.
Artificial intelligence intensifies it because intelligence can make genuinely useful centralization much more attractive. A shared system may see across institutions that currently fail to see one another. It may preserve memory across staff turnover, translate technical languages, coordinate supply, model consequences, and find routes through administrative systems that currently leave human beings carrying the entire burden between them.
That is an extraordinary capability.
It creates obligations. It does not come with a crown.
¢¢¢¢
Durán has found one unusually legible version of the future. His subjects are worth taking seriously precisely because some of them are willing to describe elite technological rule as desirable rather than hiding the ambition behind neutral institutional language. His forthcoming book argues that these ideas have moved from fringe political theory into the orbit of people with enormous capital, technology, and political influence. (Simon & Schuster)
Defeating that project would be good.
Then, the harder work begins.
The Nerd Reich is the easy case because it likes the crown.
The difficult case just works.
It predicts well. It coordinates beautifully. People adopt it voluntarily. Public services improve. Each new dependency has a defensible reason.
Nobody abolishes democracy.
Nobody declares himself king.
Nobody needs to.
The question is whether we can keep the intelligence, the coordination, the competence, and the protection while refusing to let the institution carrying them own the conditions of its own correction.
Tomorrow, unfortunately for everyone hoping I had become an anarchist, we begin by admitting that somebody still has to coordinate all this shit.