It’s Not Too Late To Stop The Machine
Artificial intelligence is more than just a technology. It’s another method of control in the toolbelt of the elite.
In 2023, Italian philosopher Matteo Pasquinelli released a groundbreaking re-appraisal of the history of artificial intelligence titled ‘The Eye of the Master’. Taking its name from Karl Marx’s seminal study ‘Capital’, the “eye of the master” refers to the watchful eye of the factory master during Britain’s industrial age, supervising the division of labour through a disciplinary regime that controlled the time, space and relationships of the worker. As a ‘social history of artificial intelligence’, Pasquinelli offers up a new epistemology that frames computation, and, in turn, AI, as less rooted necessarily in the development of computer systems and “intelligence” themselves, but a reflection of the logical intelligence required to perform acts of labour and the social processes that constitute our society. By tracing AI back to the pre-history of algorithms themselves by referring to the ancient Hindu Agnicayana ritual, where a thousand identical bricks are meticulously organised into a reconstruction of the god Prajapati, the philosopher demonstrates how humanity’s practice of ritualistic labour has led to the deep learning models we are familiar with today, like OpenAI’s GPT and Google’s Gemini. With the discovery of steam power and the jacquard loom, industrial society began dreaming up a way of life where we were free to enjoy the world as we please, with an automated labour force capable of performing the menial tasks that keep our world spinning. In his 1891 essay ‘The Soul of Man Under Socialism’, Oscar Wilde wrote that, “Man is made for something better than disturbing dirt. All work of that kind should be done by a machine.” If the machine can take the reins on the gruelling aspects of life, then the conundrum of how a socialist society can functionally operate equally is solved. The so-called Godfather of AI, Sir Geoffrey Hinton, takes a similar line of thought. When asked by ITV’s Robert Peston how the increased productivity brought on by AI can be received equitably, he responded simply with, “Socialism.” This does not come solely from an excitement for the advent of artificial intelligence but the ways that those who hold the technology will come to use it.
These concerns were just as pertinent in the Industrial age, where the Machinery Question ruled economic debate. How do these new technologies impact our social classes and our economic development? What AI offers is potentially the most dramatic of reorganising the labour force in human history, making it more efficient and even more extractive. As Pasquinelli describes, “the machine is built by the division of labour in order to achieve a more accurate calculation and extraction of surplus value.” Developments in algorithmic machinery stimulate analysis of the division of labour of the time, an analysis that alienates the worker from the knowledge that went into crafting these new tools. Marx wrote in his manuscript ‘Grundrisse’, or ‘general social intellect’, “This road is dissection— through the division of labour, which gradually transforms the workers’ operations into more and more mechanical ones, so that at a certain point a mechanism can step into their places.” The changes to work have already begun, with LinkedIn estimating that 70% of the skill used in most jobs will change as a result of AI by 2030.
The term ‘artificial intelligence’ emerged in 1956 at the Dartmouth Workshop thanks to computer scientist John McCarthy. McCarthy later expressed regret for this, believing that people are too inclined to misinterpret the term and overestimate what this new school of cybernetics and automata could do. This we know to be true. The connotations of AI signal a level of mysticism which has come to define much of the work that Silicon Valley now puts out. AI is an ideology purveyed by a cabal of technical elites who hope that their training models can one day become autonomous and replace humanity. This ideology keeps most of us at arm’s length, allowing the technocrats to run rampant with our data in the pursuit of a world of unheard-of wealth and freedom. But this ideology also holds little regard for what happens we actually get there. Some believe that a world free of labour will cause levels of depression, alcoholism, social withdrawal and even early mortality to heights we are not prepared for. There is also the chance that wealth inequality— which has already grown 50% in the last eight years, according to a London School of Economics report— could soar beyond comprehension. “There is no social safety net generous enough to protect people from the mental health effect, and society from the political turmoil, that would follow from such widespread disappointment and alienation,” wrote Eric Posner, Professor of Law at University of Chicago Law School. With very little impactful oversight yet in place tokeep the development of artificial intelligence in check, the mysticism is ideal as it allows to continue using humanity and its data as a petri dish for our replacement.
But let’s say that there is merit to the mysticism. Speaking to Doug Youvan, a biophysicist who writes on the societal impacts of AI, he wasn’t totally convinced. “I would challenge the premise that AI should be demystified entirely,” he wrote over email, “However, I distinguish between mystical and dogmatic. The former invites humility and awe; the latter enforces belief without question. A healthy public understanding of AI should preserve the mystical— acknowledging the complexity and unknowns— while avoiding dogma.” It is true that more awe-inspired reaction to AI has the potential to instil a culture of curiosity and of interdependence. Except, it feels too often like AI is treated as an independent force that is being allowed to stay the course no matter what despite its destructive possibility. I think there are benefits to the mysticism, they just work in the favour of the tech oligopoly. Much of the public do not understand the work that Google, Microsoft, Alibaba and the like are actually doing. This puts us right where they want us.
As Pasquinelli sets out in his book, the roots of computation can be found in a desire to control society through processes of automation. Society has always functioned within certain controlling parameters, upheld through enforcing agents, in order to maintain social order and reprimand devious behaviours. What we are seeing, first in the hyper-connectivity brought on by smartphones and the Internet, and now in the blind embracing of AI technologies, is a restructuring in how these parameters are enforced. Social control initiatives are swiftly being commercialised using digital tools. For example, Chinese company Baidu holds lucrative contracts for the expansion of CCTV and facial recognition drones across the country. Whilst AI-driven data analytics firm Palantir— who the New York Times report have received nearly $1 billion in contracts from the Trump administration, who have tasked them with compiling cross-departmental data on the American people into a consolidated database— continue to offer a predictive policing model to nations despite evidence of its racial biases. (Palantir have also played a central role in aiding the Israeli genocide of Gaza using their predictive algorithms). These contracts reveal the reality of the transformational opportunities of AI.
It is for this reason that it has never been more vital to emphasise how technology is, above all, a social process. Technologists require a mangling of technology and magic to hide what AI means for us in 2025. As products of the social systems they are born in, technologies bear the marks of rules and values that define our society. This is particularly so in the case of AI, where the human data it relies on contains human biases. A plethora of research papers have already been conducted analysing the ways that deep learning models contain both implicit and explicit biases with regard to race, gender and sexuality. “They reveal important glimpses on the existing structure of inequalities that, through an increasing reliance on formalised algorithmic and AI techniques, could be automated,” write sociologist Laura Sartori and computer scientist Andreas Theodorou in one paper from the journal ‘Ethics and Information Technology’.
It is vital that we get AI right. Speaking to 032c, Pasquinelli expressed that, “The positive effect of AI is that we are rediscovering the capacities of our bodies to produce knowledge and cultural abstractions. However, as a result, we will have a different social composition of manual and mental labour, which means society will be organised according to new hierarchies.” The way these hierarchies manifest is yet to be seen in full but the glimpses we have been offered point to the rapid entrenchment and automation of many of the ailments already inflicting hurt on the most marginalised in society. It is now of the upmost importance that AI is used to empower rather than control, inviting fruitful collaboration across practices to quell public fears rooted in falsities and ignorance in the spirit of democracy. “The future of AI must not be shaped solely by corporate or geopolitical interests,” argues Youvan, “it should reflect humanity’s collective wisdom and ethical diversity.” AI could bring us the freedom our species has long been looking for but only if we make sure of it.