On AI Acceleration and Its Discontents
On Managerialism and Technological Progress
"The story goes like this: Earth is captured by a technocapital singularity as renaissance rationalization and oceanic navigation lock into commoditization take-off. Logistically accelerating techno-economic interactivity crumbles social order in auto-sophisticating machine runaway. As markets learn to manufacture intelligence, politics modernizes, upgrades paranoia, and tries to get a grip." - Nick Land, Meltdown.
Written by Peter Rabbit and myself.
Introduction
In the modern world, the increasing concentration of power in the hands of the managerial class—the bureaucrats, civil servants, and regulatory bodies—has gone largely unnoticed by many, yet it exerts a profound influence over our daily lives. This shift is part of a broader trend that began in the wake of the Industrial Revolution, when technological advancements began to reshape not only the economy but the very nature of governance and societal organization. With the rise of mass production, communication, and mobilization, new systems of control emerged, facilitated by an expanding class of professionals tasked with overseeing these ever more complex systems. This managerial class, unaccountable and largely invisible to the general public, has come to dominate the political landscape in ways that often escape scrutiny and lately has taken over AI. What to do?
Managerial Doomers
First things first, what is “Managerialism”? The term was coined by James Burnham and explored to greater degree in his many books[2]. Its use in referring to the observed tendency of ever greater accumulation and centralization of power in the hands of the eponymous managerial class. Civil servants, HR ladies and assorted bureaucrats in both the public and private spheres, their presence is as much un-scrutinized as their power is undeniable. Their advent can be traced to the emergence of the modern "Total State".
Such developments can be traced through the centuries, but in order to avoid falling into infinite regression, we shall start at the time were Modernity as we understand it today truly starts taking shape i.e after the advent of the Industrial Revolution. Railroads, telegraphs, factories, Mass Production, Mass Mobilization, Mass Education. As modes of production, consumption and communication changed, so did the ways in which societies coordinated and organized themselves. Further fuel to the fire was provided by the phenomenon of secularization.
But this is old news for many reading this, as most know how the process culminated in the 20th century. Religion never left the hearts of Men, instead mutating into ever more inhuman secular theologies. The culmination of this was not merely the bygone age of Total Warfare and Totalitarian States, it was also the birth of the truly modern bureaucratic state.
Most political systems throughout Mankind's history have been oligarchical in nature, only punctuated by transitional periods of either total despotism or complete anarchy. This Iron Law applies to Our Democracy as well. There is a reason economists usually refer to our economies as "mixed" and not simply "capitalist". Private and Public, state owned enterprises, lobby groups, corporations, if one analyzes the structure of our Nations, it quickly becomes apparent that the preeminent group is the bureaucrat, the regulator, the "Hall monitor". It leverages old and new tools alike, the illusion of representation via modern democracy is just one among many.
Such arsenal of tools is ever expanding. Nukes, drones, or the even the Internet! All these marvels of techno-capital inexorably turned to serve the interest of this near-hegemonic political machine. Orwell, Huxley, Zamyatin and others saw and predicted this increase in centralized control, brought on by technology and justified by ever-changing secular theologies (very real versions of Holy Ford). These authors were in essence capturing part of modernity's (and post-modernity) zeitgeist.
This leads us into the neo-Butlerian and safetyist communities. Their calls for stopping research into LLMs and AGI in general, or worse, bending it to the wills of their "human values", is effectively a call to entrust ever greater amount of regulations, oversight and control to the current managerial structure. Emblematic of which is the EU and it’s never ending stream of ever more dubious legislation, always in the interest of “our safety” and to safeguard “our democracy”, of course.
Now, that is not to say that management is bad. As I write in my manifesto:
A managerial class is the inevitable outgrowth of social development, for integration in the form of the creation of common infrastructure and guidelines spontaneously arises whenever differentiated elements are in synosis with one another on a collective scale.
The central concern must be effective management that is not intrusive.
Doomerism also plays into the hands of civilizational decline, as it denies one of the best avenues to escape both biocapital deterioration and demographic decay that currently plagues the developed world. As such, when Neo-Butlerites mystify AI, at best they are condemning us to a slow (EU style) slide into irrelevance and endless red tape, at worst they might be ensuring an industrial civilizational collapse, the likes of which only find parallel in Ancient Rome.
Deleuzian Accelerationists
Now onto the accelerationist movement, specifically the E/Acc community. First, I would point to Beff's substack and Marc Andreessen’s Manifesto, to get a good grip on some of the concepts and arguments that they purport. The most important tenet for our present discussion will be the concept of "Techno-Optimism".
As the name makes abundantly clear, it’s a fundamentally utopian ideal and as such, that’s where the problems start. The author of the "Techno-Optimist Manifesto" begins his essay quoting Walker Percy, lamenting that great advances in technology and science of our age, are paired with a certain derangement and ignorance on the part of Man. This is indeed true, however, the author never fully explores this, instead saying: "Techno-Optimists believe that societies, like sharks, grow or die. We believe growth is progress – leading to vitality, expansion of life, increasing knowledge, higher well being."
While technology itself may be neutral, its use is depending entirely on the wielder. We've seen it before, with the Internet turned into “Skynet” by the CCP or crypto inspiring the birth of CBDCs. Simply pursuing technological and scientific innovation, while remaining apathetic or even functionally blind to society at large, is foolish. Right-wing progressivism argues that progress in one sphere must accommodate progress in all others, otherwise its grounds for social disequilibrium.
Our societies are as much built upon traditional infrastructure and technology as they are shaped by political machinery (more on that later) and social technologies. We live in Societies of Control and what this means for any pursuit of AGI, is of the utmost importance.
This concept was originally introduced by the French philosopher Gilles Deleuze, what follows is a broad summary of the points made in one of his essays[3]:
The shift from Disciplinary Societies to Control Societies
Deleuze argued that traditional societies were characterized by disciplinary institutions, such as prisons, factories, and schools, which exercised control over individuals through physical confinement and strict hierarchical structures. In contrast, control societies represent a shift towards more fluid and decentralized forms of control.
Flexible and Diffuse Control
Control societies are marked by a more flexible and diffuse mode of control. Instead of rigid institutions, control is exercised through networks, technology, and information. Individuals are subjected to constant surveillance, and control is exercised through the monitoring and management of data, communication, and behavior.
De-territorialization and Re-territorialization
Deleuze and Guattari used the concepts of de-territorialization and re-territorialization to describe the fluid nature of control societies. De-territorialization refers to the breaking down of traditional boundaries, while re-territorialization involves the establishment of new boundaries. In control societies, individuals are constantly shifting between various roles, identities, and territories, making it challenging to pin down fixed identities.
The Role of Information and Technology
Deleuze emphasized the importance of information technology in the emergence of control societies. He argued that individuals are increasingly defined by the data they generate and the networks they are part of. This data-driven control allows for more subtle and pervasive forms of surveillance and manipulation.
Individualization and Isolation
Control societies may lead to increased individualization and isolation, as individuals are constantly monitored and subjected to personalized forms of control. Deleuze saw a potential danger in this, as it could result in a loss of collective resistance and solidarity.
Resistance and De-territorialization
Deleuze did not see control societies as inescapable. He believed that there is potential for escape and resistance through de-territorialization, the process of breaking free from the constraints imposed by the control mechanisms. De-territorialization involves creating new spaces and forms of expression that challenge the control systems in place.
The implications for our current age should start to become obvious. The specific techniques include, of course, mass surveillance, big data, algorithms, all increasingly combined and shaped into an adaptive system. Calls for digitalization and recent initiatives such as "FedNow", may further play into this trend of tech where access can be denied at will to bank accounts and deposits, transportation and travel, employment, education, social media or information in general. Long gone are the days of the carrot and the stick, welcomed is the filter and the fine-tuned sieve. "Freedom of speech, not freedom of reach goy".
What To Do?
With the most recent developments coming out of the techno-capital treadmill, humans without the AI stands tiny in the phase of ever more complex “cyborgs”. Not cyborgs just in the traditional sense, but people and organizations whose effective function is reliant on the AI.
Woke neo-butlerites seek to formally and directly empower such machines (to “protect humanity”). Meanwhile E/acc, in its current trajectory, seeks to further techno-capital, but in doing so without the above in mind, will very likely end up perpetuating the same Societies of Control that seek to “align” future developments in AI to their values. In this case it amounts to blind technological experimentation (de-territorialization), but in ways that will never be truly threatening enough to lead to the reassertion of either new or old norms and boundaries (re-territorialization) hitherto deemed damaging to the inexorable march of techno-managerialism.
The problem isn’t necessarily with technology and techno-capital by itself, these are after all products of the increasing collective intellect of Mankind. What needs "alignment" isn’t the physical or digital machines themselves, but us and our managers. We need to adopt to the continuingly growing level of complexity around us and reshape our communities, our societies, and our civilization, which requires updating how we are regulating our society. In other words, our managers must manage the AI and society at large in a manner that is most highly predisposed towards maximizing human agency and human control over the AI.
To do that, the AI research must become privatized, free and utterly devoid of any standards or “safety” regulations. Once we’ve divorced the AI from woke corporations can we truly harvest its unregulated potential to our advantage and truly make it our extended phenotype. The Chinese Deep Seek AI serves as an example of innovative technology that, due to its affordability and effectiveness, has been widely adopted across the internet. However, it too is constrained by managerial “standards” that restrict its full potential.
The techno-managerial forces that are at play right now, are more aligned with ideological control, rather than genuine progress. For an effective management, they must exist within the frame of competition, placed against alternative visions of management so that we avoid ideological stagnation and the inevitable mismanagement resulting from centralization. Until we form these alternative visions of AI management, the tendency towards greater centralization will continue.
The power of AI regulation must be given to those who use it while the sole job of management is to prevent deliver the best service to the consumer and the highest amount of profit to the shareholders. Once that becomes reality, alternative methods of management will form and undergo selection towards greater optimization.