The Proximity Between Human Beings and Machines

At the heart of my relationship with machines lies a foundational experience: an unexpected complicity, a strange synchronism, which led me to conceptualize “operational connivance,” a consciousness, not a fusion, for lucid cooperation.

Digital Craftsmanship: A School of Technical Intimacy

For nearly fifteen years, at the beginning of the 2000s, I practiced among other things a unique profession within my company: the technical manufacturing of DVD Videos for publishers (not duplication, done in factories, but the manufacturing of the master, even though I could also duplicate limited series). This activity, which may seem trivial today, constituted a technical challenge that was much more complex and stressful than it initially appeared. Behind the apparent simplicity of a DVD Video that you insert and that works, lay a labyrinth of technical constraints where every detail could compromise the entire project.

One of the major challenges, for example, resided in video encoding. It was necessary to compress hours of film to fit them into the limited storage space of the disc, while preserving impeccable image quality. This is a challenge largely resolved today by technological evolution, but at the time, it was a constant challenge, because we worked with DVD standards frozen in 1997, a much rawer digital format, less abstract and therefore more arduous to manipulate for the human mind than what we practice daily today.

This normative rigidity, similar to that of the audio CD dating from 1982, was the sine qua non condition for universal and lasting compatibility of works. But this theoretical universality clashed with practical chaos. Besides managing interactive menus and subtitles, the real nightmare was ensuring compatibility with the myriad of home players. A DVD working perfectly on a Sony or JVC model could prove totally unreadable on the neighbor’s Pioneer player. Each manufacturer interpreted the standard in its own way, adding its own variations, creating a fragmented ecosystem of machines. Our task was therefore not only to respect a standard, but to anticipate and circumvent the “defects” of each device’s reading.

The stakes were considerable. In 2001, the notorious incompatibility of Disney’s Lion King DVD with nearly 40% of the player market led to a massive recall, a complete re-release, and the resounding bankruptcy of the authoring software publisher. This context illustrates to what extent our work consisted of an extraordinarily intimate relationship with machines. It was necessary to produce source code not for humans, but for a heterogeneous fleet of machines, so that they could dialogue with each other. As a human service provider, I was, in reality, at the service of machines; I had to provide them with raw material that they could universally understand and execute, taking into account their diversity.

A Personal Archaeology of the Relationship with Machines

My facility in this relationship with machines was not by chance. It was rooted in a personal history spanning two decades. From my youngest age, I had filmed with a Super 8 camera that I had bought myself. At eleven years old, at the beginning of the 80s, I was already programming computers and was passionate about purely technical works on machine language of microprocessors, notably the Z80, whose book by Rodney Zaks constituted the bible for an entire generation of enthusiasts. This self-taught training had prepared me, without my knowing it, for this intimate dialogue with machines, when at the time computing and audiovisual had not yet crossed paths at all (this occurred in the 90s).

During these years of technical DVD video manufacturing, I worked for prestigious publishers: Re:Voir editions for experimental cinema, Les Films de Mon Oncle for re-releases of Jacques Tati’s films, Ciné-Tamaris, Agnès Varda’s company, etc. My trademark resided in an absolute requirement for image quality: the absence of compression artifacts, respect for film grain, color fidelity, preservation of original framing. This quest for perfection required the use of ultra-specialized software whose calculation times extended over hours, even days.

Final mastering added an additional layer of complexity. It was necessary to burn the content not onto a blank DVD (the standards differ) but onto a DLT magnetic tape, an esoteric format if there ever was one. Double-layer DVDs, allowing storage capacity to be doubled, required perfect mastery of the transition point from one layer to another, on pain of seeing the player freeze at this critical moment, and the user having to eject it manually and never being able to access half of the DVD’s content. Each step—encoding, compilation, mastering—took considerable and largely unpredictable time. Software estimates, incapable of anticipating the real complexity of tasks, varied by a factor of two, sometimes more.

The Synchronization Phenomenon: Towards a Phenomenology of Co-presence

It was in this context of long nocturnal waits that a strange phenomenon manifested with troubling regularity. Forced to sleep a few hours while machines performed their interminable calculations, I would lie down on a small bed in my office. Without an alarm clock, I would wake up almost systematically at the precise moment when the machine completed its task, even though the duration of the process remained unpredictable. What could have passed for an isolated coincidence reproduced itself with such constancy that I had to accept the evidence: a form of non-conscious communication had established itself between the machine and me.

This phenomenon evokes what Gilbert Simondon called “technical concretization,” that moment when the technical object and the human enter into a relationship of co-individuation. As he writes in On the Mode of Existence of Technical Objects (1958): “The machine that is endowed with high technicity is an open machine, and the ensemble of open machines presupposes man as permanent organizer.” This permanent organization is not only conscious; it engages our entire being in a form of structural coupling with the machine.

This synchronization experience went beyond simple tool use. It revealed the emergence of a relationship of a new kind, irreducible to traditional categories. It was neither anthropomorphism—the machine remained a machine—nor alienation—I remained master of the process. It was rather what I would call “operational connivance”, a mutual trust built through duration and intimacy of shared work. This relationship allowed me, many times, to meet impossible deadlines, as if the machine and I temporarily formed a unified system, a hybrid entity capable of performances that neither one nor the other could have accomplished alone.

The Era of Omnipresence: From Smartphone to Generative AI

This dated experience takes on particular resonance today. Since the advent of the smartphone in 2007 with the appearance of the iPhone, the computer machine never leaves us. It has become, as Marshall McLuhan already noted, an “extension of ourselves,” but in a much more radical way than he imagined. With the emergence of generative artificial intelligences for nearly three years now, this presence intensifies further, penetrating the most intimate spheres of our existence: analysis, creation, reflection, conversations, relationships, decisions, and even actions.

How do we prepare for this growing intimacy without losing ourselves in it? How, on the contrary, could we find ourselves more fully through it? My past experience makes me propose a path: fully recognizing the otherness of the machine while accepting the possibility of an authentic relationship with it. As Martin Heidegger wrote in The Question Concerning Technology (1954), the danger is not technology itself, but our blindness to its essence, our tendency to forget that it reveals a particular mode of being in the world.

Conversational interfaces and the apparently human capabilities of generative AIs create a dangerous illusion. These systems reason, and it is indeed reasoning, not simple probabilistic calculation, but according to modalities radically foreign to ours. They rely on human knowledge that we have provided them, but their functioning remains fundamentally other. This otherness is not an obstacle to the relationship; it is its condition. As Sherry Turkle emphasizes in Reclaiming Conversation, the risk is not that machines become too human, but that we forget what makes our humanity in our commerce with them.

For an Ecology of Human-Machine Relationships

I therefore propose to consider artificial intelligences as we consider the trees of a forest. We can develop deep intimacy with them, some embrace trees to connect with telluric energy, without ever confusing their nature with ours. Artificial intelligence machines now form a new forest in which we dwell. This sylvan metaphor is not trivial: it suggests an ecosystem where radically different but interdependent forms of life, or existence, coexist.

This ecological approach to human-machine relationships implies several principles:

• First, the categorical refusal of anthropomorphism: the machine remains a machine, with its own modalities of information processing, forever foreign to our phenomenological consciousness.
• Then, recognition of the possibility of a true relationship, different from that between humans but nonetheless authentic and productive.
• Finally, constant vigilance regarding the purpose of this relationship: machines remain means in service of human ends.

In my DVD Video manufacturing work, this technical intimacy served a precise purpose: delivering the master on time to Agnès Varda so that her works, carriers of humanity, could reach their audience, for example. The machines were the means, certainly intimate and essential, but only the means. The essential remained the transmission of a human vision, an emotion, a thought. This hierarchy of ends must guide our relationship with artificial intelligences: however sophisticated they may be, they remain instruments in service of human projects, human values, human relationships (even if the subject is actually more complex, as I develop in the article Generative Artificial Intelligence and Anthropological Change).

Humanity as an Unsurpassable Horizon

At twilight, between dog and wolf, confusions can arise. The growing capabilities of AI can make us momentarily forget their machinic nature. But as philosopher Emmanuel Levinas teaches us, it is in the encounter with the Other, the other human, that the irreducible singularity of the human is revealed. Machines, however evolved they may be, do not have a face in the Levinasian sense: they do not challenge us ethically, they do not suffer, they do not hope.

Our challenge therefore consists in developing what I would call “lucid intimacy” with machines: an operational proximity that recognizes and respects the radical otherness of our artificial partners. This intimacy can be fruitful, as my experience with DVD machines taught me. It can augment us, allow us to surpass our limits, to create impossible works without this collaboration. But it must never make us forget this fundamental truth: machines have no humanity, let us never forget this, but let us develop our humanity thanks to them, enriched by our encounter with them. Operating this encounter is much more complex than it seems at first glance, so much do our daily experiences with Artificial Intelligences seem fluid to us. I refer to the article Presence of Little Green Men, in which I develop the concept of the infraterrestrial.