Once we became upright bipedal animals, humans found themselves exposed and therefore in a state of fundamental need and deficiency. However, with our hands now free and our eyes fixed on the horizon instead of the ground, we gradually evolved into handy creatures with foresight. Since then, human beings have invented roofs to keep them dry, fire to prepare their meals and weapons to eliminate their enemies. This genesis of man does not only tell us about the never-ending struggle for protection and survival, but more fundamentally about our nature as technical beings, that we are artificial by nature. From the early cave drawings, all the way to the typewriter, touchscreens, and algorithmic autocorrections, technics was there, and is here, to support us in our wondering and reasoning. Everything we see and everywhere we live is co-invented by technics, including ourselves. (This paragraph was co-authored by a human.)
Examining AI through the metaphor of the deficient animal opens a fundamentally different analysis of our relationship to artificial intelligence than the more prominent tool narratives. Rather than viewing AI as a mere instrument, if we consider humans as beings with inherent deficiencies compared to other animals, AI becomes less of a neutral enhancement and more of a prosthesis: a necessary yet accidental technical supplement which makes us (in)capable to do things we couldn’t do before. From this perspective, generative AI is not merely an artificial addition but an extension and distribution of our (lack of) capabilities. Following this metaphor, we should not fear replacement of superintelligence but instead critically examine the entanglement and co-evolution between different forms of human and artificial intelligence.
While the prevailing dichotomy in western culture draws a firm line between nature and technics, that means between what is “original” and what is “artificial” (see metaphor 1), there are alternative intellectual currents. Over the last century, several philosophical movements have contributed to the deconstruction of classical metaphysics, moving away from essentialism and (the implied) hierarchical oppositions in the great chain of being. Many have gravitated towards existentialism, portraying humans as fundamentally finite entities characterized by indeterminacy or even a kind of nothingness.
Akin to this intellectual current, a strand of discourse, known today as Philosophical Anthropology, emerged around the 1920s in Germany, led by thinkers such as Max Scheler, Helmuth Plessner, and Arnold Gehlen. This school of thought intersected existential perspectives with biological observations, highlighting the inherent nudity and deficiencies that characterize humans compared to other species.
Gehlen, a prominent figure in this school, described humans as a Mängelwesen, which translates to something as “being of deficiency”. This concept hinges on the notion that humans, through their innate vulnerabilities, find realization and fulfillment through constructing artificial environments, essentially becoming beings steeped in technology. We lack sharp teeth, claws and thick fur and thus developed clothing, tools and weapons to compensate for our biological shortcomings. Accordingly, these thinkers dispute the idea of artificial surroundings and culture as a “second nature,” arguing instead that our technological habitat is not just a primary, but our only nature. They flip the narrative: technology is not simply a tool for human purposes, something that only mimics nature and is therefore always secondary to nature. They emphasized that humanity is fundamentally artificial, thereby altering the equation where nature is the origin and artifacts as derivatives. On the contrary, technics is crucial to human constitution, which is thus artificial by nature.
Consequently, technology isn't merely an external tool invented and deployed by great minds for specific purposes; it is a fundamental part of human evolution. Humans have only transcended animal instincts through technology. More recently, philosophers like Bernard Stiegler continue on the premises of the German Philosophical Anthropology by suggesting that humans and their “essential” qualities—such as anticipatory thinking and reasoning—are essentially co-shaped by technology at every juncture in evolution. For example, the invention of writing not only reshaped the process of evolution through an externalized technical memory, but also paved the way for universal reasoning as seen in mathematics. In other words, technology co-constituted mathematics as a universal language of the human being; it is not an essential characteristic of our nature as animal rationale.
Thus, humans are prosthetic beings without essential characteristics, which means they are defined by an originary default (Défaut d’origine). The human does not only invent, but is co-invented by his technical supplements as well. This prosthesis should therefore not be understood merely as an “artificial body part or limb” that restores a natural quality but rather, quite literally, as something placed in front of us (pro-thesis). Prostheses are externalized organs positioned before us, forming new co-dependencies with us and co-shaping our evolutionary path.
The emergence of humans was an accidental genetic event, a point with which a Darwinian would agree. However, according to Stiegler, Darwinists underestimate the role that technics have played in anthropogenesis after a certain point in history. Since we became bipedal creatures and freed our hands to grasp tools such as flint, pottery, and weapons, technics have co-shaped our biological evolution.
This prosthetic nature should not be seen as something merely negative and compensating for a lack. While the prosthetic analogy may initially signify a deficit or handicap, the German Philosophical Anthropology and Stiegler argue that this negativity is not only something bad and undesirable. The indeterminacy of human beings is also what gives it their endless positivity; to determine itself freely, to form a civilization, to sublime drift into art, etc. Because we are nothing by nature, because we lack any substantial being, we are free to become anything, to invent or transcend ourselves (see upcoming transhumanism metaphor), over and over again.
However, the flip side is that this task remains perpetually unfinished. In the tradition of modern Western philosophy (from Hegel to Marx and Nietzsche), one could still speak of alienation, thereby assuming an origin from which we have become estranged and to which we should return. However, in this “artificial by nature” paradigm, any hope of returning, of restoring our society according to our human nature or how life is supposed to be based on the natural order of things, has the tendency to become meaningless. We will never arrive at our human destination, fulfill the plan of God, or complete the destiny of ourselves as authentic human beings, because we lack any such origin. But how does this inherent artificiality in the human condition configure our relationship with AI?
By viewing AI as a prosthesis instead of a tool, machine or robot, and ourselves as a deficient animal instead of a Demiurge, artisan or autonomous subject, a thorough examination of the intertwined development of life and technics becomes indispensable. Here, the idea proposed by one of Stieglers teachers Derrida regarding technics as a “pharmakon”—a term denoting technics as both a remedy and a poison—offers a valuable lens through which we can scrutinize contemporary AI. Technology wields the dual potential to heal and harm humans or society, as evidenced throughout history.
Although it remains imperative to acknowledge that technology has propelled humans to transcend the animal kingdom, it also harbors the latent ability to foster regression. For instance, while innovations like the printing press and the rise of (affordable) newspapers and books have significantly enhanced literacy among the general populace, the rise of the digital attention economy can strain our co-evolved cognitive abilities, now leading to problems in terms of literacy. Similarly, advancements in transportation, from the invention of the automobile to the development of high-speed trains, have revolutionized human mobility and connectivity. However, these same advancements have also contributed to urban sprawl, environmental degradation, and a sedentary lifestyle, posing new challenges to both our physical health and the planet.
Nevertheless, starting from the fundamental entanglement and co-evolution between humans and technology, the above examples of regression cannot simply mean deviating from an origin and corrupting the human essence. We cannot simply get rid of tech and then restore our “natural” cognitive qualities, “natural” ecosystems and “natural” number of daily steps as we deconstruct human nature. On the contrary, we must be more precise in defining what we mean by harm and potential in relation to the human, which has become more challenging as we can no longer fall back on essentialism. Accordingly, in the field of AI this metaphor helps us move beyond not only the naive doom-thinking and overly optimistic reflections that dominate today's AI debate but also the commonplace banal assertion that every new technological innovation carries both risks and opportunities. Be more specific!
So, what does this concretely imply for generative AI? After bypassing these extreme positions and ethical clichés, we can zoom in on specific use cases and assemblages of humans and technology to arrive at a more nuanced understanding. For example, consider how writing has co-evolved with the advent of new writing technologies, with ChatGPT being the latest advancement, as recently examined by FreedomLab Fellow Leonardo Werner.
As he rightly argues, like all innovations within the Technological System, as defined by Jacques Ellul, ChatGPT embodies a relentless drive for efficiency and optimization. However, it also holds the potential to revolutionize our writing capabilities in positive ways. For instance, it can enhance our ability to structure texts, correct grammar and spelling errors, streamline the writing process, and potentially develop new co-writing skills that enable instead of disable creativity and imagination, as some writers are currently exploring.
However, good writing encompasses many more, harder-to-define qualities than creativity (which has become a clichéd soft quality) and harder-to-define risks than optimization. As many have noted, a fundamental harm of generative AI might be the tyranny of the average: the reduction of our vocabulary and possibly even our expressiveness. Or as Leonardo mentions, the true quality of writing perhaps lies in anything that resists metrication.
Beyond that, we have to admit that what we (who is “we” to begin with) consider good writing is neither universal nor ahistorical but will also change in this process of co-dependency and co-invention. This could go two opposite ways: after the initial wow effect, we might increasingly start to recognize AI-generated writing as banal, full of clichés, one-dimensional, high-school level and good means to deviate from it, or we might increasingly, perhaps also unconsciously, begin to accept it as the new standard or norm for how good texts should be written and what we expect from others. Will we start to avoid using the words “delve” and “intricate interplay” or are these and other LLM-standards becoming the new (hidden) norms?
This, of course, is closely tied to the open question of whether the entire globe will rely on just a handful of LLMs or if we will co-evolve in a world of specialized and niche AI models.
The impact of ChatGPT and similar technologies will undoubtedly vary across different social classes, cultures, levels of education, and more. The open philosophical question is then how fundamental this variation truly will be. In the realm of international business, we might witness the emergence of a new “AI Globish” as the standard, characterized by a proliferation of autogenerated titles and “deep dive into” jargon. This trend is likely to extend to social media, dominated by marketeers and influencers under significant pressure to contribute to the never-ending AI content stream. In journalism, there will likely be increased efforts to establish and implement ethical frameworks and professional codes, perhaps also more space for language specialization. In literature, given the current economic landscape, we might see a sharp divide between AI-written or enhanced books on e-commerce platforms, that means high volumes of ebooks sold for minimal profit, and a small group of writers distinguished by their idiosyncratic and authentic styles. Groene Amsterdammer has recently revealed that over 4% of content on certain e-commerce platforms is already autogenerated. You can test it yourself by typing in “Oppenheimer biography” on bol.com; you need to scroll some time to find a book written by a human author.
Beyond writing, similar trends can be expected in other forms of content creation, such as video and music. However, beyond content creation, Generative AI will also lead to many unforeseen combinations of human and technical AI supplements. In this context, we might be looking, or searching, in the wrong direction. For example, many now claim helpful AI agents are the next big frontier. However, I am skeptical about AI agents autonomously planning our holidays and scheduling our calendars becoming the major feature of next-gen generative AI, contrary to Altman's claims. The idea strikes me as an uninteresting and anthropomorphic view of AI technical supplementation, akin to the early belief that personal computers were only good for calculations and spreadsheet management.
As series like Altered Carbon, Black Mirror and the movie Her profoundly demonstrate, it may be far more intriguing to explore the general topics of love and intimate relationships rather than just functionality and automated assistance. If we are able to develop helpful AI agents, I am inclined to believe they will be more “suited” for caring, grieving, and emotional and sexual intercourse rather than for planning, scheduling, and making holiday lists.
Nevertheless, even these entanglements remain very anthropocentric. Perhaps the new technical supplements are more akin to Alien Intelligence, making AI agents more like Alien Agents instead of Samantha. Or perhaps more like the enigmatic and unconventional forms of intelligence in the realm of being such as fungi and viruses, rather than a personalized Scarlett Johansson. However, we will explore that idea further when we move beyond these human-centric metaphors in the next sections, where we explore the organic and mythical metaphors around AI.
(This article was co-authored by AI.)
