It’s hard to imagine what was going through the mind of Enoch Taylor, a British blacksmith and inventor in the early nineteenth century, when some workers, inspired by the ideas of the legendary Captain Ludd, began to destroy the cropping machines he had created with hammers he had produced. The Luddite motto “Enoch made them, Enoch shall break them!” echoed throughout England, stirring the conscience of the Victorian bourgeoisie and landed gentry, still shaken by the memory of the French Revolution. As the anarchist philosopher John Zerzan argued, the Luddite movement was a complex phenomenon and it cannot be reduced to a single demand or battle. But it can be essentially seen as workers’ existential struggle to defend the dignity of their work. Luddism was not only an expression of the fears and socio-economic distress of an exploited social class. It was also, and above all, a cause of alarm and anguish for the nascent English bourgeoisie. Their power was violently challenged by a moderately organised popular movement that made clear demands regarding the working conditions of the nascent proletarian class. For one of the first times in the modern era, the movement brutally depicted the nature of individuals as agents faced with the mechanisation and automation of manual labour. Today, we know how that historical experience ended. But the minds of the protagonists, both wealthy and proletarian, must have been full of anxiety and apprehension at the time. The anxieties of the Luddites about a future that could, at least in their aspirations, be constructed through their social and political action. The apprehension of the wealthy about a future in which their prerogatives and privileges were challenged by a violent group seen as the wretched of the Earth.
It would be interesting to know if Elon Musk, or other driverless car tycoons, experienced similar mental states last year when a group of self-described “Luddites”, with the advantage of evening twilight during the Californian autumn, managed to destroy self-driving taxis belonging to the Cruise company in San Francisco. The vehicles were rendered “blind” by placing a traffic cone on the bonnet, which prevented sensors on the roof of the car from “seeing” the road. The result was a mad herd of mechanical unicorns taking over the streets of San Francisco until they inevitably crashed into a wall or another vehicle. The strategy employed in these acts of vandalism, though far removed from the hammering violence of General Ludd’s original militias, betray a similar logic of action, which belies the idea of Luddism as a technophobic reactionary movement. The original Luddites used the technology at their disposal (the hammers produced by Enoch) to reclaim the dignity of their work. Similarly, the new Californian Luddites exploited the technical knowledge at their disposal (the functionality and positioning of the sensors produced by Cruise) to make their protest against an urban planning policy that favours the experiments of Silicon Valley companies over the needs of citizens for well-functioning public transport, green mobility incentives, and a vision of the city as being less dependent on cars. Just as the logic of the British Luddites emphasised the capacity of the exploited to act when faced with the new phenomenon of work automation, the logic of their modern Californian counterparts is based on asserting the independence and multiple cognitive capabilities of individuals faced with their automation. As noted by the philosopher Matteo Pasquinelli in his book The Eye of the Master: A Social History of AI, the current development of artificial intelligence for performing low-skilled jobs – that have traditionally been seen as requiring a low level of specialisation – actually hides a high level of cognitive sophistication. Even for the automation of actions that are second nature to us after a little practice, billions of dollars have been spent, and countless engineers have lost sleep, only to see the product of their efforts crash into a wall because of a traffic cone placed on the bonnet of the car.
In other words, as Pasquinelli says, all work is cognitive work. Yet, there is an aspect of cognitive work that we fail to automate with artificial intelligence. In the 1990s, the American philosopher Hubert Dreyfus, in trying to explain Heidegger’s concept of ‘instrument’ (Zeug), used as an example the experience of driving a car. An expert driver, Dreyfus argued, does not have a clear and constant awareness of the vehicle he is driving: his actions are dictated by habit and, in particular, by his ontological state as an agent within a network of possible actions and meanings that are present and immediate in theact of driving (what Heidegger calls Zuhandenheit). The analytical gaze that reflects and examines the act of driving (Vorhandenheit), on the other hand, is not immediately present as the driver acts. That is the result of the subject distancing from the network of possible and meaningful actions from which the concrete action occurs. In other words, when a person drives, he performs the act of driving, not the act of thinking about driving. Despite these observations, current AI technologies for driverless vehicles, and other tasks, do not function like drivers immersed in the network of possible and meaningful actions with their cars. Rather, they act like observers on the roofs of cars, measuring with extreme precision the distance from other cars and the width of the road. The two actions are intrinsically different and, as such, will lead to different results.
What does this observation tell us about the state of work under the present technological regime? As already noted, the cognitive work performed by machines, in the current state of the technology, is essentially different from human cognitive work. A driver in a car with a traffic cone planted on the bonnet could stick his head out of the window to look at the road or manoeuvre the car by using the rear-view mirror or turning his head. Or, obviously, he could stop the vehicle if he thought the conditions were too risky for safe driving. These actions, as we saw in San Francisco, are not possible for the artificial intelligence systems developed by Cruise. This indicates that fears (or hopes) about the disappearance of work in the context of automation are unfounded. Just as the need for workers did not disappear after the adoption of machinery during the Industrial Revolution, it is not possible to replace human cognitive labour with a series of automated actions that are, essentially, different. What might happen is the disappearance of those repetitive and mechanical jobs that can be automated without losing important aspects related to human cognition, jobs like data entry which requires the employee to transfer data from one format to another. Such employees, which in any case constitute a minority of the workforce in developed countries, could be replaced by machines, but, in this case, we would be facing the same phenomenon that we saw during the Industrial Revolution.
Image: Photo by Discover Savsat on Unsplash
