Relational Intelligence

Hello, 2026

Living, thinking, working, and having sex with machines: today, we are bound to technologies more intimately than perhaps ever before. This is not a particularly original claim. In fact, it has been demonstrated over and again across multiple empirical studies. Sixty-five percent of 16- to 39-year-olds in Germany, for instance, discuss mental health concerns with chatbots and report experiencing “closeness and a sense of connection” in doing so.1 And yet, stubborn misconceptions persist—especially in academic settings—about how we, as human beings, ought to relate to this new reality. There are real anxieties circulating in universities too: might generative artificial intelligence (AI) in some cases already outperform professors at conveying knowledge? In this short essay, I want to take up that question, not in terms of the comparative abilities of humans and machines, but in terms of the kinds of intelligence involved.

Histories of Wounding

Following the cosmological wound inflicted by Copernicus (we are not the center of the universe), the biological wound dealt by Darwin (we descend from early hominids), and the psychological wound opened by the discovery of the unconscious (there is no free will, only drives), the human being stands there, its naive narcissism in tatters.2 Add to this that humanity is currently dismantling the very conditions of its own existence on this planet, and the picture darkens further. From this I draw a first conclusion: intelligence adequate to our time will be an intelligence of humility. What especially agitates our wounded selves right now is the growing awareness that, through the wholesale delegation of decisions to artificial intelligence, we are ever less in charge of our own inner household.

The fear of our progressive fusion with AI, however, deserves something other than alarm; it calls for a kind of loving attentiveness. After all, this encounter with AI is also, in some measure, an encounter with ourselves as its co-creators. New ideas are taking shape, meanwhile, about what a liveable coexistence with technology might look like. Donna Haraway calls this principle “making kin”: cultivating kinship even where no natural kinship is apparent.3 My second conclusion follows: intelligence adequate to our time will be relational intelligence, or, in other words, an intelligence of connection to other people and technologies alike.

Human Beings Need New Images

The dominant metaphor for AI remains, persistently, that of the tool: the machine figures as a useful, manageable, and mute “thing”. I read this devaluation primarily as a reflex of anxiety. AI is not a simple tool like a hammer. It is more like a dialogic interlocutor, one that listens, accompanies, and to which we sometimes entrust our most intimate secrets. Yet precisely here lies a danger. Wherever relationships form, closeness and insight are always accompanied by new possibilities for dependency, projection, and manipulation. This requires us to conjure new images, perhaps the image of a resonating body that mirrors us and, in so doing, transforms us. Or the image of a paradoxical, not uncritical bond, one into which we find ourselves newly entangled each day. Such images can open the conceptual space for thinking about ourselves and AI. They ask of us a willingness to admit other perspectives and self-understandings, to resist the demand for immediate certainty, and to persist in unknowing. To return to the question of intelligence, those who frame generative AI as nothing more than a tool or an assistance system imagine themselves standing at a safe distance. Those who understand AI as a dialogic actor move beyond that posture and in doing so practice relational intelligence.

In Close Contact

Relational intelligence between human beings and AI can, at its best, give rise to a form of relational knowledge. This knowledge emerges, for example, through sustained dialogic contact with a chatbot, something we are currently researching at the Berlin University of the Arts. In this process, the human being is no longer considered a fully sovereign author. For this to be possible, AI would need to be understood—technically and scientifically—far more thoroughly than it currently is, and far more subject to meaningful control. Rather we, along with our thinking, knowing, and feeling, are continuously being co-produced by our technologies. The philosopher Karen Barad captures this mode of entanglement with the concept of “intra-action”.4 Only when we take this connection seriously does the deeper dimension of our relationship to contemporary technology begin to reveal itself gradually and in its full complexity. To move at this level means, above all, learning to navigate uncertainties and blind spots, both our own and those of technologies.

This returns us to the question of the kind of intelligence required. This must be a dialogic intelligence, the capacity to remain in contact with what is strange, contradictory, and unresolved, without reflexively reaching for control, either in relation to technology, in relation to knowledge, or in relation to oneself. Perhaps this is precisely where the deepest challenge of our present moment lies. To close this deliberately affirmative essay on a sober note, I’ll give the last word to Fran Lebowitz: “I am more worried about human intelligence than I am worried about artificial intelligence.”

Enno Schramm is a researcher and theorist at Berlin University of the Arts (UdK), where he investigates relational knowledge emerging from human-AI dialogue. His work sits at the intersection of philosophy, technology studies, and contemporary art practice. He is particularly interested in how new forms of intelligence and understanding develop through engagement with algorithmic systems, and what this means for human creativity, autonomy, and connection.

The original article was published on the Culture Shifts Substack.