Will Artificial Intelligence Dismantle the Managerial Class?

AI poses the first real challenge to managerial society, but the cure may be worse than the disease.

Managerialism refers to a society in which managers - bureaucrats and technocrats - become the de facto ruling class of a society. In today’s political discussions, managerialism is typically associated with the work of James Burnham, who wrote the book The Managerial Revolution in 1942. But an even earlier conception was presented in Bruno Rizzi’s 1939 book The Bureaucratization of the World. I summarized the managerial revolution in a previous deep read.

Newsletter #63: Understanding the Managerial Revolution

Interestingly, both Rizzi and Burnham were Marxists by heritage. Rizzi’s book is about how in the Soviet Union the bureaucracy had replaced the aristocracy and bourgeoisie as the exploitative class. So while references to Burnham and the managerial revolution are a mainstay on the right today, managerialism was originally a left wing concern, and indeed there’s a long history of leftist critique of bureaucracy.

The problem for critics of managerialism is that managerialism has been inevitable under the conditions of modern society. Managerialism fundamentally arose because of the revolution of size and scope. The dawn of the railroads and the second industrial revolution created the first large scale organizations. Prior to around 1835, there were only a handful of organizations that employed even more than 100 people.

As large organizations grew that involved vast geographic scope, vertical integration, complex technology, and large numbers of employees, the techniques of management needed to be developed to run these huge organizations, whether that be the Pennsylvania Railroad, Standard Oil, or General Motors. Industrialization created gigantic cities like New York and Chicago that required large amounts of infrastructure and management to function. Big business and big cities led naturally to big government, a necessity to stave off oligarchic control of the country.

The managers eventually became a class that across all these organizations effectively controlled the country and reaped outsized gains. This wasn’t just for large organizations either. The Washington, DC urbanist Richard Layman used to say of local government there that, “Big government remakes little government in its own image.” The same is true for other organizations. The managerial revolution and the managerial mindset penetrated even small organizations like churches. I previously hosted a podcast with Chris Schlect about the managerial revolution in American churches.

So long as we have these large organizations, getting rid of managerialism has been impossible. That’s why, for example, one of the fiercest critics of managerialism, the paleoconservative writer Sam Francis, came to the conclusion that the American “soft” managerial regime could only plausibly be replaced by a “hard” managerial one (translation: fascism or something like it).

Enter artificial intelligence. For the first time, AI offers the possibility of managing large scale, complex entities without a correspondently large and independent managerial class.

Let’s be clear, we are talking about the future. But given what we have already seen AI demonstrate, there’s every reason to believe it is possible at some point.

Even basic algorithmic management is already facilitating this shift, a transition to what I previously labeled “impersonal” (as opposed to bureaucratic) authority. Consider Uber. Uber drivers don’t report to a boss that assigns work. They are directed by an algorithm. True AI, such as driverless car technology, then replaces the driver altogether.

Managerialism is replaced when this same process is applied to managers. AI will start to replace managers and technicians by taking over the basic administration functions. AI also, critically, might “Uberize” the remaining managers and technicians, subjecting them to impersonal, algorithmic control from the senior most levels of an organization.

While the exact details are of course hazy, and the future cannot be predicted, AI looks to be a centralizing technology. It may well restore, validate, and extend the power of senior executives over organizations by eliminating or disempowering managers.

Imagine a world in which much of the US government bureaucracy has reached an advanced stage of impersonal AI management. This alone would restore immense power to the elected President and White House. Future Presidents would have far fewer problems with career civil servants attempting to sabotage or subvert their agendas.

When you think about this more broadly, it’s easy to come to the conclusion that the cure (AI) may be worse than the disease (managerialism).

For example, we already face increasing challenges in speaking to an actual human customer service agent when dealing with many companies. There may come a day when it’s simply impossible, that we are completely subjected to algorithmic relationships. We may not even know what entity actually stands on the other side of that interface.

This isn’t just theoretical. Investigative reports out of Atlanta, which is one of the biggest markets for large-scale investor-owned housing, found that some tenants didn’t actually know - and couldn’t figure out - who their actual landlord was. Interactions were basically through a collection of outsourced service providers, and almost entirely mediated through technology. Hence tenants found it impossible to get problems fixed (yet could still be easily evicted if they failed to pay the rent, since like many red states, the deck is stacked in favor of slumlords in Georgia).

AI isn’t required for this to happen, but AI enables it at a scale and extent that would otherwise be difficult or impossible.

China’s surveillance apparatus, social credit score system, and full digitalization of society also shows how these technologies allow vast social control over a population.

The “AI doomers” worry about a Skynet-like system that might kill or enslave humanity, but the most likely negative outcome is AI excessively empowering dictators or an oligarchic class.

If you think billionaires or imperial Presidents wield too much power today, wait until our fully AI enabled world.

Additionally, one corollary of this kind of AI driven centralization is that it might lead to additional geographic centralization of culture and the economy. Much has been made of the so-called “rise of the rest” in terms of growing technology and startup ecosystems around the country, or people moving away from the coastal elite cities. But the core AI companies are in San Francisco.

Cloud technology and open source were decentralizing forces, enabling lots of cities to create ecosystems of software-as-a-service companies. But it may well be that ecosystems built around putting wrappers on AI engines will not be viable. This would hurt secondary and tertiary technology clusters.

It’s possible that a fully AI driven world is one in which a handful of cities really are command nodes of the overall country and economy. The tendency has already been for the highest value functions to get pulled further up the geographic hierarchy.

All of this is not a guaranteed outcome. And the possibility of it doesn’t mean that we should attempt to suppress AI technology. Had someone seen the downsides of managerialism and put a halt to the second industrial revolution, most of us would still be subsistence farmers. The gains far outweighed the losses. It may well be the same with AI. I’m personally leaning into AI and am looking forward to seeing what it can do.

But I can tell you this: get ready. The age of impersonal authority is already upon us. Will an AI powered post-managerial America be far behind?

Cover image credit: Sam Altman of OpenAI by TechCrunch - CC BY 2.0.