Extended integrity

We are cyborgs: this is an anthropological observation, not a metaphor. Our integrity as persons is no longer contained within the limits of the biological body; it extends into our machines, our data, and now into the artificial intelligences with which we think. I propose the concept of extended integrity to name this condition, and to equip the questions we need to ask ourselves in order to live within it more fully.

The test of forgetting

Imagine you forgot your phone at home this morning. When I ask this question to the groups I work with, the answers are almost always the same. “I wouldn’t have been able to get back home, all my tickets are on my phone.” “If I lose my phone, I have no way to contact anyone, I don’t know any number by heart anymore.” “Before, we used to leave for the day and tell ourselves, if there’s a problem, we’ll figure it out. Now, it’s anxiety.” And the worst, having it stolen: everything is in there, and we can no longer access it.

These are remarks made by women working in education and therapy, perfectly clear-sighted about their relationship to technology. What they describe resembles a condition more than an addiction. Inside the phone, there are transit tickets, the bank card, contacts, passwords, the calendar, emails, photos, intimate conversations, itineraries, proofs of identity. Without the phone, what is felt goes beyond the absence of a forgotten tool. Something of oneself has been removed.

What integrity means

Integrity, in the philosophical sense, means being whole, undivided. The integrity of a person is their completeness as a subject. An assault on integrity, in law as in morality, is one of the gravest harms one can inflict on someone.

We spontaneously conceive of this integrity as bound to the body: physical integrity of an intact body, psychic integrity of a coherent mind. But what becomes of it when part of my memory, of my connections, of my capacities, of my social identity, is housed in an object that is not my body, and when this object is connected to networks and processed by algorithms that escape me?

I propose to name extended integrity the condition in which our completeness as persons exceeds the limits of the biological body to extend into our technical objects and digital environments. This extension is in no way a pathology, nor a degradation of a “true” integrity that would belong to the body alone. It is an anthropological transformation of what it means to be a whole person. Our integrity lies on this side of and beyond our body. We are ourselves within the machines.

Philosophical antecedents

In 1998, the philosophers Andy Clark and David Chalmers published an article entitled The Extended Mind, in which they argue that human cognition is not limited to the brain but extends into the tools we use. Their example is that of a man with Alzheimer’s who keeps his information in a notebook: this notebook, they say, is part of his mind in the same way as his biological memory. If the notebook performs the same function as memory, it is an extension of it.

Bernard Stiegler had developed earlier, within a different conceptual framework, the notion of tooled memory. For him, technical objects are “tertiary retentions”, external supports of memory that prolong and transform our biological memory. The book is a tertiary retention. The camera is another. The phone is a third one, more complex, because it does not only preserve information: it organises our relationships, our transactions, our social identity.

Donna Haraway, as early as 1985, in her Cyborg Manifesto, proposed to think the hybridisation between the human and the machine not as a dystopia but as a political opportunity. For her, the cyborg is a figure of emancipation: by blurring the boundaries between nature and culture, between biological and mechanical, it opens possibilities that traditional categories had closed. Haraway was not yet writing about smartphones, but her proposition resonates with our current situation.

From the side of social psychology, Russell Belk had formulated in 1988, in an article that has become a reference (Possessions and the Extended Self), the idea that our possessions are constitutive of our identity. The objects we keep, the ones we pass on, are part of the person we are for others and for ourselves. Belk was writing before the smartphone era, but his thesis finds in this object an exemplary case: the phone is today the technical condensation of nearly all our personal extensions.

More recently, in 2011, the psychologists Betsy Sparrow, Jenny Liu and Daniel Wegner published in Science a famous article, Google Effects on Memory, which shows that when we know an information is easily available online, we memorise less the information itself and more the place where to retrieve it. Human memory has reorganised itself around the availability of the search engine. The empirical finding meets what Stiegler and Clark had foreseen by philosophical means.

Katherine Hayles, in How We Became Posthuman (1999), had raised a related question starting from cybernetics: what happens when information ceases to be thought of as attached to a body, and becomes an autonomous quantity that circulates between human and machinic supports? Hayles does not arrive at dematerialisation; on the contrary, she reminds us that we are bodies thinking within technical devices, and that no information circulates without material support. Her reading avoids the transhumanist fantasy without conceding anything on the anthropological transformation under way.

I myself proposed, in an earlier text entitled The intricated person, a related notion to describe what is happening between us and artificial intelligences. The intricated person names the cognitive coupling between a person and the AI with which they think. The concept of extended integrity, which I propose here, does not contradict the intricated person: it broadens its perimeter. The intricated person speaks of what happens when we think with an AI. Extended integrity speaks of what happens at the scale of the whole person, of their identity, of their informational possessions, of their connections, of their social presence. The phone that was stolen from me harms my cognition, and far beyond that: it is my whole person that is attacked in its extension.

The fusion of the physiological self and the digital self

This fusion within us, between our physiological self and our digital self, is already here. It is neither good nor bad, it is what is. The important word is “already”: this is a present, not an anticipation. A present that is not even twenty years old in its current form, since the smartphone as we know it dates from 2007.

Serge Tisseron, in Le jour où mon robot m’aimera (2015), raises a question that has stayed with me ever since: “The question is not whether we will communicate with avatars as with human beings, but whether we will reduce the humans behind their avatars to those avatars, and also why we so often communicate with the human beings around us as we do with avatars, by reducing them to their appearance or function.” The risk is not that we treat machines as humans. It is that we treat humans as machines.

This is what can happen if we keep believing that the integrity of a person is limited to their biological body. Everything that is in their machines, from data to conversations, would then count only as exploitable matter. The “real” person would be the body in front of us, and all the rest, mere scenery. Yet the person is also within their machines, and respecting their integrity requires respecting this extension of themselves.

The intimate singularity

There is much talk of the singularity, the concept popularised by Ray Kurzweil to designate the hypothetical moment when artificial intelligence would surpass human intelligence. It is often pictured as a distant and abstract event.

I believe the singularity is already here, but that it is intimate. It manifests itself in the daily life of every person, in this imperceptible fusion between the biological self and the digital self. Finding one’s way on a phone, speaking to one’s children through a phone, having one’s spelling corrected by an AI, being guided in one’s choices by recommendation algorithms: the singularity inscribes itself there, in the fabric of ordinary life.

Mark Alizart, in Informatique céleste (2017), recalls that nature contains a code, DNA, which drives cell division and which meets memory and the surrounding environment. He suggests that the machines we build may not be opposed to the biological, but seek to refabricate it. Artificial neural networks function in a non-linear way, in the image of our own neurons. If nature is already a form of computing, artificial intelligence prolongs nature rather than breaking with it.

This hypothesis helps to think the intimate singularity. If the machines we build prolong natural processes, extended integrity is not an anomaly. The living has always practiced this extension: prolonging itself through tools, modifying its environment, inhabiting it. The beaver builds a dam that is part of its vital ecosystem. We build smartphones that are part of our cognitive and relational ecosystem. The difference of scale is considerable; the principle is no doubt the same.

Thinking with machines

At the moment I am writing this text, I am speaking into a dictation device. A piece of software then transcribes my speech, I reread it, I reorganise it, I entrust it to an artificial intelligence so that it proposes a layout and references I would not have thought of on my own. I reread its proposal, I cross out, I rewrite, I get my hands back into the work. The text that comes out is intended for humans, and bears my name, and I keep claiming it as my own writing. This writing is a process, and each tool I use within this process modifies what I am able to think. Ideas are not already whole inside my head, waiting to be deposited on the page: they form and transform themselves in the back-and-forth between speech, transcription, reading, the machine and rereading.

Marshall McLuhan formulated this point as early as Understanding Media (1964): the medium is the message. The tools through which we communicate shape our way of thinking at least as much as what we communicate. Walter Ong extended this idea in Orality and Literacy (1982) by showing how the invention of writing transformed our mental operations: without the written trace, we cannot think in the same way as with it, because oral speech is entirely dependent on living memory and on the context of exchange. Writing by hand, writing on a typewriter, writing on a computer, writing by speaking into a dictation device, writing with an artificial intelligence that proposes reformulations: each of these is a different mode of thought, which makes different things appear.

I notice for instance that my oral thinking, spoken into the dictation device while walking, is more fluid, more associative, more emotional than my written thinking. The thinking I elaborate in dialogue with an AI is more structured, better referenced, but it is also exposed to the risk of an averaged language that I must correct so that it becomes mine again. Each of these forms brings something the others do not, and all of them participate in the person I have become.

Extended integrity, understood in this way, is not only a passive fact we would undergo. It is also a know-how to acquire: an attention to the way each tool orients our thinking, and a responsibility in the choice of the tools we adopt or set aside. Thinking with machines requires knowing how each one orients thought, and cultivating this knowledge as one used to cultivate, in another era, the art of rhetoric or of conversation.

What we have to cultivate

Intelligence has become, as the AI manufacturers themselves say, a “commodity” in the same sense as water or electricity, that is to say something everyone can benefit from. If this is the case, intelligence is no longer what defines us as humans.

Yuval Noah Harari, in Sapiens (2015), suggests that what differentiated homo sapiens from other human species was the capacity for imagination: fabricating narratives that bring communities together, believing in things that do not exist materially (gods, nations, money, borders). His thesis is that it is narrative capacity, even more than cognitive capacity, that made humanity.

This line of thought interests me, and I would like to extend it differently. What seems to me to remain ours, and even more so what we have to cultivate as ours, is shared and sensible experience: attention to the other, and the fact of having planned a programme only to find that nothing happens as planned, because we are together in a unique group and things transform between us that were not programmed.

Olivier Houdé, in his work on cognitive resistance, shows that human learning passes through the capacity to resist one’s own automatic thinking. This resistance demands effort, trust, an environment of safety. Machines do not resist themselves, they optimise. They face neither doubt nor the fear of being wrong in front of others. This confrontation with shared uncertainty holds something properly human.

But I would like to avoid here a formulation that would be backward-looking. If I simply say that these human qualities are to be preserved intact in the face of machines, I miss my subject. Within the condition of extended integrity, these qualities are not natural reserves left untouched by the digital. They are themselves transformed, traversed, put to the test by our fusion with machines. My attention to the person in front of me is now lived in competition with the notifications on my phone, which I know can reach me at any moment. Sensible experience shared with a group, going out, laughing, walking, falling silent, is traversed by the awareness that all of this can be recorded, photographed, shared. And the uncertainty we explore together in deliberation is silently modified by the possibility of asking an AI for a probable scenario before we have even taken the time to search.

The stake, then, is not to preserve a human-core that would remain sheltered from the digital. It is to learn to cultivate human qualities within the cyborg condition itself, knowing that the tools we use modify those qualities. Cultivating attention to the other today demands a precise know-how: knowing the moments when one must put down the phone, and those when its presence is neutral. Cultivating shared sensible experience demands knowing when the image stands in the way of experience, and when it prolongs it. Cultivating a thinking that remains my own demands learning where, in the course of my reasoning, I can call on an artificial intelligence, and where it is essential to keep searching alone or with other humans, so that thought forms itself in resistance and not in the comfort of immediate suggestion.

Ethical consequences

Taking the concept of extended integrity seriously has several practical consequences :

• First for personal data. If my data are part of my integrity, their non-consensual exploitation is not a mere commercial inconvenience but an assault on my person. Debates about privacy gain from being reformulated in these terms: protecting persons in their extended completeness, more than protecting abstract information. Knowing the various artificial intelligences, knowing which ones collect data mandatorily and which do so optionally, is not a secondary technical knowledge: it is a knowledge of self-protection.
• Then for education. Children and young people also have an integrity that extends into their machines. Policies that aim to separate them from it, banning phones at school or social media before fifteen, need to be thought with greater nuance. I am not saying that all is well, nor that children can do whatever they want with screens. I am saying that separating a child from their phone touches their extended integrity, and that this calls for accompaniment, not for a brutal prohibition.
• Lastly for mediation, whether cultural, educational or therapeutic. If the persons we work with have an extended integrity, their digital dimension is part of the relationship, not a side aspect or a distraction from a “real” relationship between bodies. Their digital life is part of them. Their images, their conversations with AIs, their practices on social networks are dimensions of their existence to be welcomed and accompanied, rather than interpreted from the outside as symptoms.

Cultivating humanity in the cyborg condition

Nostalgia is tempting: a time when integrity would have been simple, contained within a body uncomplicated by machines. Such a time may never have existed. The book was already an extension of memory; the family photo album an externalised support of identity. What changes with the digital is the scale, the speed and the permanence of this extension. The principle itself is ancient.

Naive enthusiasm is just as problematic: believing that machines are going to augment us, to make us better. Stiegler, with the concept of pharmakon, reminds us that every technical object can be both remedy and poison. The smartphone that connects me to the world can also isolate me from the person in front of me. The artificial intelligence that helps me write can also prevent me from thinking for myself. The use is what decides, and the consciousness we bring to inhabiting this condition.

Tim Ingold, in Lines: A Brief History (2007) and in his work on the anthropology of perception, proposes the notion of “shared inquiry”: living together an experience of knowledge that builds itself in the long time of shared practice, and that cannot be substituted by any finished product. Ingold writes from a world in which machines have remained external to his inquiry. Mine is not quite the same, because machines are already part of the inquiry. When I work with groups, artificial intelligence is in the room, in everyone’s phones, in the automatic transcriptions we read together, in the images we generate to think with. Shared inquiry now extends to the tools we mobilise. I do not refute Ingold; I augment his thought through the conditions of the present.

Our responsibility, in this condition, is to learn to inhabit consciously our extended integrity. This requires knowing what is in the machines, knowing what we deposit in them, knowing how those deposits come back to us. It also requires cultivating human qualities, attention, connection, presence, within the fusion with machines, rather than apart from this fusion. The cyborg condition can then become the occasion of a more conscious humanity, provided we work it as such.

This is a philosophical, educational and political project. Philosophy here does not consist in giving answers, but in equipping our apparatus of questions, to take up a phrase I propose to my students and to the groups I accompany. What questions are we asking ourselves today about what we have become? What questions must we learn to ask in order to act better tomorrow? Extended integrity is one of these tools of questioning. It does not say how to live. It names what there is to live, and allows us to ask what we want to do with it.