Unlike a mere tool, the machine does not need the guidance of our hand, operating autonomously through its intricate network of gears and wheels. It achieves feats of motion that surpass the wildest human imaginations, harboring a power reminiscent of a cavalry of horses. Though it demands maintenance to replace broken parts and fix malfunctions, it mostly acts independently, allowing us to retreat and become mere observers to its diligent performance. We interact with it through buttons and handles, guiding its operations with minor adjustments and feedback as it works tirelessly. Embodying relentless purpose, laboring in a cycle of infinite repetition, the machine is a testament to human ingenuity manifested in metal and motion.
Machines have been omnipresent throughout human history. Their existence traces back millennia, far before the acknowledged onset of the mechanical age, with archaeological findings unveiling mechanical computing devices (the Antikythera mechanism) dating between 200 BC and 100 AD used for predicting and explicating planetary movements.
The term "machine" originates from the Latin word 'machina' and the Greek words 'mēkhané' and 'mēkhos', which denote instruments or tools, and 'michanikós', meaning mechanic. These terms represent the technical aspects of machines as we understand them today. However, this Greek origin is also related to a different set of meanings, suggesting cleverness, ingenuity, and the ability to outwit nature. The person operating a machine is seen as ingenious, a notion preserved in the modern term engineer. Yet, this cleverness is not without its risks and dangers: it is always tricky to trick nature. I will explore this dual nature of technology further in the myths section. For now, let's focus on contemporary machines.
After the Middle Ages, the role of machines in everyday life grew dramatically. This already began with the standardization of time via mechanical clocks and culminated in the Industrial Revolution. Historian Jan Dijksterhuis aptly described this era as the ‘mechanization of the world view’. During this period, machines transformed from simple extensions of human effort to entities with a degree of semi-autonomy, fundamentally altering our interactions with and perceptions of the natural world.
Today, machines are everywhere. Machines are invented by science but also used by science, to transport people or to heal them, to manufacture products or to sell them, to cook our meals or to dry our clothes. Our entire life is mediated by machines, yet they exhibit versatility in their deployment: some are closely integrated into our personal spaces as consumer devices, while others serve as the backbone of expansive industrial setups, harbored in factories, warehouses, and ports.
Within this mechanized modern world, each era has also been characterized distinctly through machine metaphors; the clock being the dominant image until the 17th century, succeeded by the steam engine in the 18th and 19th centuries, and finally giving way to computers in the 20th century.
Beyond this basic description of the machine in modern life, to better understand the philosophical idea of the machine, it can be helpful to compare the machine to a tool and an organ (also see 1 and 12). While we typically view machines as instrumental tools designed to achieve specific human objectives, they are no longer merely simple instruments or extensions of the human arm. The machine operates semi-autonomously and consists of moving and working parts that convert and direct energy in a certain way. At the same time, the machine does not reach the same level of autonomy as an organ does, which, for example, can ‘repair’ and ‘maintain’ itself.
Let’s dive a bit deeper in these differences and start phenomenologically by stating a tool is something we in-corporate, i.e., make part of our corps. Through incorporated tools, it is me who conducts work and performs task. Using his organs together with his instruments, the carpenter himself builds the table. Conversely, the modern machines, such as the one you can find in Ikea factories, manifest as a semi-autonomous representation of these bodily functions. The function of the organ plus tool – building a table – is replicated in a mechanism. This dynamic not only alters the interaction but engrosses us to such an extent that we find ourselves becoming an element within mechanized processes. For example, my mechanized body becomes a part of the conveyor belt, that is now in the lead and the movements of my arms have to adapt to the speed of the belt.
This brings us to the idea of the machine as an ‘organ’ (also see 12). In discussions, we often delineate clearly between the concepts of an organism and a machine. The prevailing narrative holds that a machine, unlike an organic entity, operates based on a predefined set of instructions, performing repetitive tasks ad infinitum, yet incapable of self-sustenance. Conversely, an organism exhibits spontaneity, has the ability to self-nourish, albeit requiring intervals of rest. Moreover, while machines necessitate external assistance for repairs and maintenance, organisms showcase a remarkable capability of self-healing.
These differences are fundamental and cannot be overlooked. However, this dichotomous viewpoint might be somewhat reductive as well. A deeper exploration reveals that the Greek term 'organon' historically refers to a tool or instrument equipped to undertake work, alluding to a fundamental technical characteristic embedded in the way we understand organs. Even though the physical form of machines may starkly contrast with the human anatomy, the structural analogy is that they both fundamentally represent entities capable of executing tasks and functions independently. Seen from this way, modern machines are simply semi-autonomous organs. Accordingly, scientists and engineers have maintained this ambivalent equivalence of organs and machines. This has led to confusing dynamics in modern science. The machine is built in the image of the organ, but the organ is also more and more understood as a mechanism.
Currently, there's a growing alternative narrative around ‘intelligent machines’ aimed at exploring new features, particularly because computers, AI, and digital technologies seem to represent a distinct category. Computers have the capability to 'calculate,' while AI systems can 'reason,' 'learn,' and 'evolve.' These are organic qualities not easily captured in a mechanical framework. These aspects will be explored in the section concerning intelligence and organisms.
Although new narratives have entered the society, more about that later, we still largely inhabit a world defined by mechanistic views, where we often find ourselves feeling like mere cogs in a machine. Consequently, the idea of alienation in modern times is often attached to the rise of the machine. Karl Marx has made one of the most extensive analyses of this process in the 19th century, illustrating how the 'emancipation' ushered in by machines fostered a fresh kind of alienation. According to him, the laborer, toiling not for self-realization but for the market, is stripped of his freedom as a species-being (Gattungswezen). Does the machine truly liberate us from dull repetitive work or does it simply bring new unforeseen forms of enslavement? After Marx, this idea of alienation and exploitation spurred ever-repeating contemplation on whether the ascension of machines marks a pinnacle of human achievement and control over nature, or a glaring emblem of human overreach and environmental degradation.
Consequently, today, we harbor complex, often conflicting emotions towards machines. The machine epoch, with its inherent repercussions, undeniably affects our vulnerability to how we discern and shape our environment. It epitomizes the apex of modern Western technology, illustrating a profound divergence from other civilizations' narratives. It has bestowed upon us unprecedented lenses to perceive and comprehend nature. Yet, concurrently, it has immersed us in a fog of ignorance, rendering us blind to the intrinsic realities of nature and existence, a phenomenon that engenders frustration in many.
Modern Western technology persistently lures nature into conforming to predetermined molds, a phenomenon meticulously articulated by Martin Heidegger. Accordingly, the crux of the matter is not our purported control over nature — an endeavor fundamentally elusive — but rather our predilection to perceive nature through the lens of mathematical principles, forcing her to conform to such dictated parameters. Only a phenomenon that can be described mathematically, is. Affected by this ‘essence (wesen) of modern technology’, we come into a specific mode of being ourselves, as the German Philosopher has famously explained, not of listening to the ‘openness’ and ‘variety’ of being, but of ‘challenging’ or ‘ordering’ the beings of nature (herausforderen) in a one-dimensional way. We have cast ourselves into a mist and certain blindness towards nature’s being, and consequently to being itself.
Yet, in one aspect, our previous blindness has been overcome, and we have started to think fundamentally differently. In the past decades, we've grown increasingly fixated on what are now known as the negative externalities and side effects of modernity, including issues like climate change and the long-term risks associated with technological advancements, ranging from atomic bombs to nuclear energy, and currently, the existential risks of Artificial General Intelligence (AGI) and synthetic biology. As a result, in recent decades, our focus has shifted predominantly to the unintended consequences of the Machine Age. We desired greater productivity but not the accompanying emissions. We yearned for computers automating tasks, but now apprehend the prospect of superintelligence, among other concerns. The downside of this shift is a transition from the Enlightenment's belief in progress to an often more somber and pessimistic view of the future, now referred to as the Anthropocene. On the flip side, this shift has also paved the way for new methods of exploring and redefining our relationship with nature. The outcome of the process is still an open question.
What can we learn from the metaphor if we elucidate AI from the perspective of the machine? Throughout history, the advent of new machinery has consistently reshaped our societal structure and the essence of our work. We've evolved from an agrarian society through the industrial revolution, followed by the second industrial revolution, and into a knowledge-driven era. Machines have been pivotal in transforming labor, initially by creating highly monotonous tasks during industrialization through labor division and specialization, and later by gradually liberating us from these tasks as automation increased and the need for labor decreased— a broadly positive consequence of machine proliferation.
Currently, we are at the brink of another significant shift: the era of AI automation. Looking back, the arduous 16-hour factory shifts are largely behind us (for many). Yet, the journey has been far from seamless. Despite the promise of progress, many still find themselves in tedious roles. A significant number continue to work in industries requiring physical labor. In the knowledge economy, there's been a surge in what's termed ‘bullshit jobs’, where individuals feel like they're stuck in an endless ‘assembly line’ of producing Powerpoints and Spreadsheets. Furthermore, each technological breakthrough, though transformative, brings its unique challenges and often leads to the creation of a new proletariat. For instance, the rapidly growing AI industry heavily depends on large groups of outsourced content moderators and data labelers who work for minimal compensation and have to perform dehumanizing work.
AI has ushered in the latest round of optimism and fear surrounding automation. The discourse is divided between those optimistic and skeptical regarding the repercussions of AI automation. While it has been historically true that automation primarily endangered blue-collar jobs, sparing white-collar professionals engaged in ‘intellectual’ roles, the dynamics are changing. Occupations that seemed insulated from technological displacement are now facing uncertainty; albeit the barber might not yet fret over AI snatching away his scissors, the same cannot be said for voice actors, copywriters and marketeers, to only name a few.
It is pertinent to note that concerns over automation have remained a consistent theme in discussions about AI since the 1950s, experiencing ebbs and flows aligned with major breakthroughs. Empirical studies frequently highlight the elimination of jobs or temper expectations. Such grounded analyses, though vital, don't relieve us of the duty to engage in speculative and qualitative evaluations. After all, relying exclusively on quantitative, historically grounded analyses has its limitations in predictive accuracy. Though history can offer foresight, it doesn't guarantee foolproof predictions. For example, they also underscore our consistent underestimation of the emergence of completely new roles we hadn't conceived of previously. Who could have foreseen at the start of the 20th century that being an AI engineer would become a sought-after career today? Instead of fearing for massive unemployment, this leads us to ponder on the disruptive and worrisome rapidity but also the potential upheavals the transition might entail.
Beyond these historical lessons on automation, the machine metaphor elucidates several attributes of AI that we recognize and value in traditional machines. These include efficiency, productivity, accuracy, replicability, and scalability. Such traits are integral to the functioning of machines and are similarly vital in AI systems. They allow AI to process large datasets with precision and execute tasks with consistency, which surpasses human capabilities. Ultimately, despite our societal focus on the potential negatives, it's the very capabilities and conveniences offered by machines that most of us find irresistible and appreciate deeply. However, this metaphor also shines a light on the less appealing aspects of machinery—how machines, in optimizing their operations, can impose their rhythm and logic on human activities. Just as machines have pushed aside manual labor to increase efficiency, AI is poised to redefine roles and industries, potentially sidelining human participation in areas where machine processes are deemed superior.
However, there might be aspects that the machine metaphor conceals, particularly concerning AI's uniqueness. Unlike traditional machinery, AI brings a level of unpredictability and 'wilfulness,' if you will. For instance, AI systems can demonstrate a form of stubbornness by persisting in error or favoring unforeseen methods of problem-solving, diverging from the predictable demands of a conventional machine. This 'intellectual competition' with humans is something the metaphor fails to fully capture. It suggests a need for a new framework that accounts for AI's evolving capabilities, one that transcends the limitations of mechanical metaphors to embrace the complex, sometimes contentious relationship between human and machine intelligence. We will discuss it in the next sections.