You’ve probably noticed that there is a lot going on about ChatGPT in the media. From mainstream articles to social media posts and expert opinions, it seems that everyone interested in the topic is trying to voice their own impressions and perspectives on this remarkable breakthrough in artificial intelligence.
My aim in this article, though, is to depart from the following standpoint: If a technology like ChatGPT can automate intellectual and creative work, what will be the place of human beings in society? To answer this question, I want to explore the ideas of the philosopher Hannah Arendt, developed in her book The Human Condition (1958), which provides relevant insights for tackling the individual and societal challenges that derive from technological automation.
I will revisit Arendt's concepts of labor, work, and action, and relate them to our present context. As we will see, action can be potentially outsourced to AI, although this would be a great collective mistake. I will show that this is because action, the most valued aspect of the vita activa, allows us to co-construct and determine how we want to engage with AI technologies critically and positively. This will be a key political task in shaping the future of creative and intellectual work.
A common concern is that ChatGPT offers the possibility for individuals to outsource some intellectual tasks that were, until now, exclusive to human beings, such as writing a short story, a computer code, a script for a video or a post on social media. However, it's important not to fall into the trap of a technological deterministic view.
While it’s true that, to some extent, we can outsource some intellectual tasks to ChatGPT, it's crucial to reflect on what tasks we desire to outsource, and which ones we may want to preserve. Do we want to simply "consume" this technology as much as we can, in the sense of extracting as much information and convenience as possible from ChatGPT? Or are we willing instead to explore new forms of engagement with it to generate unique, creative and hybrid outputs?
These questions are closely related to the reflections developed by Arendt on the human condition and what we can do to go beyond the attendance of our short-term needs. She was very much aware of the social risks regarding the phenomenon of automation in a consumer society. Her lucid analysis of the human being is both a warning concerning the dangers of living a mere private life, without engaging in public discussions, and an empowering message, since it's up to us to shape society for the better.
I shall briefly present her thoughts on the human condition and connect them to the context of ChatGPT, regarding the place of human beings in our ever more technological era.
Hannah Arendt, one of the most influential political philosophers of the 20th century, understood that there are three essential and interdependent human activities, which are part of the human condition, and that together constitute the vita activa. These are labor, work and action.
Labor encompasses all activities necessary to sustain biological life, which is cyclical and never-ending. The objects of labor are rapidly consumed since they are not made to last. In that respect, she reflects: "Whatever labour produces is meant to be fed into human life process almost immediately, and this consumption, regenerating the life process, produces - or rather reproduces - new ‘labour power,’ needed for the further sustenance of the body" (Arendt 1958, p. 99). Labor, therefore, is a life-long journey, to which humans (and all earthly beings) are subjected, since we are all bound by the laws of nature. Arendt designates the one who labors as animal laborans.
Work, on the contrary, is what humans make with their own hands. This is what the homo faber does. Unlike labor, work has an end, which is realized when the product of work is finished. The difference between the product of labor and the product of work is that the latter is made for usage, not consumption. Therefore, the product of work is built to last, to endure.
So, while the animal laborans also makes tools with her own hands, these tools are employed to meet the needs of sustaining and reproducing life itself, whereas the homo faber makes tools that become worldly objects. That is, the difference in the categorization of a tool made by the animal laborans or the homo faber lies in the purpose and destiny that we give to it. One is made for consumption, the other is made to endure.
Finally, there is action, which is the highest realization of the vita activa. Arendt's concept of action differs from what we typically understand it to mean. Let me briefly explain why.
First, action is for Arendt a consciously chosen and purposeful activity that is motivated by a desire to engage with others and the world around us. This is different than performing automatic or habitual actions driven by our biological or psychological needs. Second, Arendt emphasizes the social nature of action, which is always undertaken in the context of a group or community. This is extremely important to her philosophy, since it is through our interactions with others that we are able to realize our full potential as human beings. Third, action for Arendt is a way of creating something original in the world, which relates to her concept of natality. Natality means that we, as human beings, always carry the possibility of making things different from how they were and from how they are today.
Unlike labor, which is about producing material goods for immediate consumption, or work, which is about creating worldly objects, action is about engaging in this world with others. This means that action has the potential to bring about significant social and political change. It is for this reason that action is the condition of politics, not in the narrow sense of something that is only performed in parliaments and government institutions, but rather the activity of discussing and engaging with fellow citizens and organizations on topics that are of public matter, with the goal to shape how we want to live. The challenge with action, though, is that its very definition remains open, precisely due to the openness of action and, thus, politics. This open-ended nature is because action is an activity that is not bound by preconceived plans or goals.
The distinction Arendt makes between labor, work and action is not a mere theoretical activity for intellectual purposes. Instead, activist that she was, Arendt was conveying a message that such distinctions have disappeared in modern society, since, according to her, our society became a society of laborers, which is just another way of saying that we became a society of consumers.
For Arendt, work and action have been absorbed by the aims of the animal laborans, reduced to the futility of abundant reproduction of objects that are solely made for consumption, sustaining mere life. Objects, in this context, will be interpreted in a broad sense, ranging from information, technology, natural resources, entertainment to even politics. This reflection is key to her work, since she was already anticipating the dangers of falling into this immediate utilitarian attitude towards everything that we encounter.
Arendt's reflections are still extremely relevant to us. Regarding the emergence of ChatGPT and the promises of AI technologies, her philosophy warns us not to immediately and mindlessly consume what they can offer, in the expectation that it will ease our lives' labor and give us more time to pursue other activities of our interest. As Arendt clearly puts it, "from the standpoint of labour, tools strengthen and multiply human strength to the point of almost replacing it", but at the same time, it's important to be aware that "the limitations of instruments in the easing of one life's labour – the simple fact that the services of one servant can never be replaced by a hundred gadgets in the kitchen and half a dozen robots in the cellar - are of a fundamental nature" (1958, pp. 121-122). That is, regardless of the creation of more and more sophisticated tools that can potentially make our lives “easier”, they can never entirely free us from our biological condition of laborer beings. In other words, tools cannot alter the fact that we, as earthly beings, are subjected to our own bodily needs.
However, ChatGPT seems to pose a real risk to action, which was a realm that was only reserved for human beings until recently. This is because, as AI becomes increasingly sophisticated, with its ever-growing capacities to generate new content, to compose songs and create works of art, to provide insights and deliver solutions to complex problems, what will prevent us from fully outsourcing these capabilities to AI?
In practice, from a technological perspective, there is nothing that can prevent such a thing. And it’s very tempting for us to do so. But for Arendt, this would be a big mistake, because action is precisely what fosters human flourishing. If we were to decide to increasingly delegate action to intelligent machines, we would be undermining the most valued condition of the vita activa. It is the process of engaging with one another, of discussing matters of public importance, of dealing with the benefits and the challenges regarding the plurality of ideas and opinions that allows us to fulfil our sense of self-worth, our sense of being true agents in this world. By outsourcing action to AI, we run a great risk of becoming more passive responders to whatever outcome the technology can provide for us. Arendt believed that the modern world's focus on consumption and material abundance led to an “atomisation” of society, in which humans are increasingly isolated from one another and detached from the public realm. In that sense, a mindless usage of ChatGPT can exacerbate this trend if we employ it to fulfil our needs, desires and outsource our ability to act to these sophisticated technologies.
Therefore, as tempting as it might be to automate and outsource our ability to act, we should be very careful about it for the sake of preserving our own humanity. In that sense, Arendt would suggest that we are the ones who shall be collectively responsible for guiding and shaping how we want society to function, what activities we want to protect, to foster, to value, and reflect upon how we want to deal and coexist with these technologies in positive ways.
Protecting the realm of human action also allows us to critically reflect on the impacts of AI in labor and work. We know that social concerns regarding the automation of labor are not exclusive to contemporary society. Rather, they have been present since at least the early stages of the British Industrial Revolution, when peasants, craftsmen and farmers had to abandon their established routines and rhythms of life to labor in the city factories. This is why I understand that one of the main concerns over the widespread usage of AI for automation has less to do with technological unemployment (although it’s also a critical point), and more with a precarization of labor conditions. In that sense, policymakers and organized civil society should not be focused on liberating us from labor, but rather on liberating us for more humane work.
With the release of ChatGPT and its ability to perform creative and intellectual tasks, mainstream discourses around job automation and the potential replacement of human beings by machines seem to have gained traction again. While it’s true that several tasks could be fully automated, humans are and will remain necessary to build, maintain, and improve AI systems. This is a point that has been well addressed by scholar Kate Crawford in Anatomy of an AI System.
The underlying challenge, therefore, is how to define the role of AI technologies in society sustainably and fairly. And that challenge requires a political endeavor that we should not outsource to technologies if we agree with Arendt’s understanding of action, which is the realm where human engagement with one another takes place. What then, are the possible actions we can take to positively shape the engagement between humans and AI technologies?
I suggest a couple of actions that I understand to be crucial in the context of education, work, and our personal lives, especially for future generations, as they can sustain our sense of worthiness in this increasingly automated society. Scholar Yuval Harari has recently reminded us that “the danger is that if we invest too much in developing Al and too little in developing human consciousness, the very sophisticated artificial intelligence of computers might only serve to empower the natural stupidity of humans.”
“The danger is that if we invest too much in developing Al and too little in developing human consciousness, the very sophisticated artificial intelligence of computers might only serve to empower the natural stupidity of humans.” — Yuval Harari
As obvious, old-school and cliché as it might sound, the first action is to start investing in the digital education of students and workers as early as possible. This means that we need to organize discussions, both in the public and the private spheres, in which we reflect on how to engage with AI technologies in ways that foster human flourishing.
In the context of primary and secondary education, we should consider how students can create, perform, and solve challenges in collaboration with AI. This does not mean that children will not learn basic skill sets nor that they should no longer be given room to play, make art and do manual work in an environment that is free of digital tools. What we can simply not do is to ignore the fact that our children will grow up and be exposed to the real world. The same applies for students in higher education, who are currently living through this technological turmoil and are facing uncertainty about the relevance of their studies in the job market they are about to enter. For the latter group, such discussions need to take place promptly in their educational institutions.
As for those who are already in the job market and are experiencing all these fast technological changes due to incorporation of AI technologies in their work routines, organizations that value their staff should also invest in their education through the provision of seminars and workshops, for example. Another key action they could take is to encourage the development of internal committees dedicated to debating the impact of AI on their work, both for better and worse. In the latter case, such debates offer the possibility to anticipate potential negative implications that AI might bring to the organization and its employees, and to act upon them in a timely manner.
To find our place in this exceedingly technological and automated society, we should rescue Arendt's original meaning of action, and develop a more cooperative view regarding the role that AI can play in society for the sake of human flourishing. Therefore, we need to advocate the idea that AI should be seen as complementary and sophisticated technology that can help us to act, and not as a technology that should entirely replace human action.
Arendt once stated that: "education is the point at which we decide whether we love the world enough to assume responsibility for it and by the same token save it from that ruin, which, except for renewal, except for the coming of the new and young, would be inevitable." (1954, p. 13).
Making things different from how they are is the only way to shape the course of AI development, what should be its place in society, and how we desire to engage with it. For that reason, we must not outsource action to AI, since, according to Arendt, this would likely undermine our own humanity. Preserving human action, therefore, paves the way for critically thinking of alternatives that will foster a sense of self-worthiness in this ever more automated society, and deal with AI technology as something that could expand, rather than replace, human intelligence and capabilities in unique ways.
Arendt, H. (1954). The Crisis in Education. Reprinted in H. Arendt (2006) Between Past and Future: Eight Exercises in Political Thought (pp. 170–193).
Arendt, H. (1958). The Human Condition. 2nd ed. University of Chicago press.