The Dead End of Digital Conscientious Objection

Faced with the deployment of generative AI, some researchers proclaim conscientious objection. I defend instead the requirement of critical presence and empowerment through practice.

The trap of disembodied critique

Several researchers, notably a collective from Toulouse, have recently formulated a “digital conscientious objection” in response to the deployment of artificial intelligence in educational institutions. This term reflects a legitimate desire to resist the technologist imperative—the one that presents every innovation as necessarily beneficial simply because it is new. This stance is healthy in its intention; it manifests disagreement with directions largely driven by industrial interests, relayed with perhaps too little discernment by major French institutions like the Ministry of Education, perhaps too focused on staying “up to date.”

I fully share the need to build critical thinking about the use of these tools. We must indeed refuse to march “as one” toward technological solutions presented as immediate and salvific. This blind faith in technical progress is a form of modernist obscurantism that substitutes belief in the intrinsic value of novelty for critical thought. It’s a sort of “digital youthism.” Michel Serres showed in Petite Poucette (2012) how each major technical mutation transforms our relationship to the world and requires us to reconfigure our very humanity. This lucidity is indispensable. Rushing forward has no power to resolve anything, it seems to me. We believe that human beings resist change, but I am always amazed, on the contrary, by how quickly we adapt to new uses—we saw this with smartphones, for example, or QR Codes during the Covid period.

But digital conscientious objection, despite its laudable intentions, constitutes in my view a philosophical and practical dead end. I believe it rests on a fiction: that we could keep our distance from these technologies, observe them from outside, maintain a position of withdrawal that would preserve our critical integrity, by not being touched by them. This posture, however understandable in the face of rapid technical deployments, seems to me to reproduce exactly the error it claims to combat: it remains in abstraction rather than engaging in critical experimentation. And this abstraction is a nonexistent myth.

Lessons from a health crisis

During the Covid period, we experienced a situation that casts harsh light on these mechanisms at work. A solution presented as “technological” was not proposed but imposed: vaccination with an experimental product that had not had time to be validated according to usual protocols and that was sold at a premium, financed by state loans to capitalist powers. This imposition was accompanied by unprecedented enrichment of major shareholders, who were able to double their fortunes in less than two years when they were already among the wealthiest in the world.

The methods deployed perfectly illustrate what Noam Chomsky and Edward Herman described in Manufacturing Consent (1988): a massive orchestration of information, systematic discrediting of any dissident voice, fierce defense of capitalist interests by states themselves. These same institutions that contributed to the majority of human beings sinking into this scientific obscurantism are today the promoters of unbridled use of artificial intelligence. Mistrust is therefore legitimate, and we have good reasons to seek to build our critical thinking regarding anything presented as a “solution.”

But, and this is where the analogy becomes instructive, it’s not because the management of the Covid crisis was authoritarian, idiotic, and strictly in service of capitalism through shameless manipulation of information and state lies, that we should have denied the potential danger of the virus. There really was an extremely contagious virus, rapidly changing forms, about which we needed to acquire knowledge and work on prevention, individual responsibility. What was necessary was autonomy and responsibility, including for doctors. Instead, the State became the prescriber, an unprecedented fact in history, and a large part of doctors renounced their Hippocratic oath by agreeing not to treat people infected with the Covid virus (if they did, they were struck off the medical register). This freedom to prescribe was forbidden to them, and those who provided care (some did, because many treatments existed) did so by pretending to treat other diseases. Each terrain, each person being singular, there existed a multitude of possible prevention and care approaches. The lie consisted in claiming that a single product could miraculously resolve this disease, which is a scientific, medical, and epistemological lie of the lowest order, but which nevertheless drew the majority of people into its belief, so panicked by anxiety-inducing discourse that they were ready to swallow any nonsense to believe they could thus reduce their risk of dying or causing others to die. Let’s not forget that Covid “vaccines” never protected against transmission of the virus, but only against severe forms for sick people. The opposite was asserted to citizens who, in good faith, had these experimental substances—not yet authorized but possible to inject only if there were no other treatments, hence the prohibition of care by doctors and the discrediting of treatments that nevertheless worked—inoculated (under penalty of losing their jobs). Moreover, side effects were far more significant than any other vaccine.

What was needed was not to set aside the danger of the virus, but to take it into account in an informed way—that is, in diversity—and responsibly—that is, with one’s autonomy. Paolo Freire, in Pedagogy of Autonomy (1996), showed that emancipation comes through the capacity to think for oneself in recognition of uncertainty, not through withdrawal or passive obedience. The antifragility that Nassim Nicholas Taleb (2012) describes is built precisely in conscious and experimental confrontation with what can hurt us.

The illusion of separation

It’s exactly the same thing for artificial intelligence. To make ourselves believe we could live without it, as conscientious objectors suggest, is complete illusion. Even if we don’t want to use it consciously, it’s already being used for us in many cases: in processing our administrative files, in the functioning of our mobile phones, in phone applications, websites, in surveillance imposed on us without our awareness, etc.

Tim Ingold invites us in Making: Anthropology, Archaeology, Art and Architecture (2013) to consider that we are not facing techniques but within them, woven with them in a continuous process of “making.” Anyone who uses GPS is already a user of artificial intelligence. You find this very practical? Then you’re already fully participating in the phenomenon. To make ourselves believe that GPS is different from current generative AIs is only a matter of degree: artificial intelligences have been present for a long time in many areas of our lives; they are part of our existential environments.

Hannah Arendt, in The Crisis in Culture (1961), reminds us that man is “conditioned”: we live in a world made of objects and systems we have created but which, in turn, shape our conditions of existence. To claim to stand apart from AI is to misunderstand this fundamental anthropological condition. We are always-already caught in this technological tangle. The question is therefore not whether we will use it or not, because we already do, but how we will inhabit this condition, how we will take responsibility in the face of reality, not fantasy.

Inhabiting rather than fleeing

We could decide to regain our own sovereignty over our relationship to territory and not use GPS—that would be quite laudable. But perhaps we can also be more aware of how GPS works, what industries operate them, what dependence this puts us in, what we might learn from it, and how we could perhaps implement other technical solutions for navigation aid that are more virtuous in terms of common good and spatial pollution.

John Dewey’s thought in Democracy and Education (1916) enlightens us here: education is not the transmission of constituted knowledge but collective experimentation, confrontation with real problems that require transformation from us. To learn is to experience a problematic situation and to transform oneself in and through this experience. Conscientious objection, by making itself believe it stands at a distance, deprives itself precisely of this possibility of experimentation that alone permits the development of true critical consciousness. It is flight, not struggle.

Guy Debord warned in The Society of the Spectacle (1967) that “the spectacle is not a collection of images, but a social relation among people, mediated by images”. If we content ourselves with refusing AI without experiencing it creatively, we remain in the spectacle of its critique; we do not enter into a real transformation of our social and technical relations. As I have often observed in my pedagogical workshops, it is by creating with AI, by exploring its possibilities and limits in a collective and reflective framework, that an embodied understanding of its stakes develops, and capacities for action are born. Staying outside means staying ignorant of one’s “enemy.”

The path of critical empowerment

Let us therefore be not conscientious objectors but conscious users. Whether we like it or not, we are caught in this tangle. Our role is not to pretend to extract ourselves from it—that’s a deception—but to learn to untangle it so it doesn’t suffocate us. This posture requires infinitely more courage and lucidity than withdrawal: it demands informing oneself autonomously and diversely, being responsible and conscious of one’s own uses, knowing free alternatives and digital actors who have ecological values, learning new knowledge, new ways of knowing, new ways of living in this situation of mutation. Let’s take interest in new ethical approaches to manufacturing computer chips, let’s try using Open Source generative AIs that we install on our own computers and see their results, let’s also use Open Source AIs operated in data centers like Infomaniak in Switzerland (via their Euria application), which has a genuine ecological approach in this domain, for 25 years.

This presupposes an approach I have experimented with numerous times in educational and cultural contexts: starting from creation. Having concrete experience with the tool, exploring its possibilities in a pedagogical framework that allows reflexivity. Using AI to write scenarios, generate images, then asking ourselves collectively: what does this transform in our relationship to creation? What industries operate these tools? What dependencies are we creating? What are our real margins of autonomy?

What Michel Serres describes with the metaphor of the parasite (1980) applies perfectly here: AI is a parasite in the system, but the parasite is never simply negative; it creates noise that can become a source of innovation and transformation. Provided we don’t passively endure it or pretend to ignore it, but learn to compose with it. Conscientious objection condemns us, in my view, to endure from afar what it refuses to inhabit—that is, to produce the opposite effect of what it advocates.

Toward an ecology of practices

The anthropological challenge that AI represents, this mutation I call a “displaced us,” requires critical presence from us, not withdrawal. We must develop what Yves Citton calls in Mediarchy (2017) an “ecology of attention” that allows us to remain attentive to the effects of these technologies on our existential environment, while remaining engaged in their experimentation.

This position is not naive. I know very well that AIs are developed by industries whose interests are primarily financial, that their ecological impact is considerable (though it is still today a thousand times less polluting than industrial livestock farming), that their deployment participates in a worrying concentration of power. But to claim to extract ourselves through conscientious objection is to condemn ourselves to powerlessness and irresponsibility. It means leaving the field open to those who ask themselves no ethical questions, who are nothing but low-grade merchants. Let’s not give them the power.

I therefore defend a third way: neither enthusiastic and blind adoption, nor conscientious objection that keeps its distance, but engagement in critical and creative practice. Learning to untangle the tangle so it doesn’t suffocate us requires getting our hands into it, having concrete experience with these tools in collective frameworks that allow reflection. It is only in this way that we can develop the autonomy and responsibility that this major anthropological mutation demands, and that we will support the development of more ethical technologies, which is possible, if we take this political stance in our everyday actions.