Artificial Intelligence in the Service of Human Connection

Generative AIs worry us with their mimetic capacity. Yet, by freeing us from time-consuming cognitive tasks, they offer us more time and energy to cultivate what is essential: our human relationships.

An ancient fear facing a new reality

Generative artificial intelligences, in the mimicry they display, give us the impression that they will replace us. This fantasized projection about competition between artificial intelligences and human beings always surprises me, particularly since these systems became capable of generating coherent texts. Yet artificial intelligences did not emerge ex nihilo in November 2022 with ChatGPT. The term itself is very broad and poorly defined.

Optical character recognition, operational since the late 1980s, is now presented as artificial intelligence. Voice recognition, which has existed for about thirty years, receives the same label as soon as we add a few recent algorithms, even though its fundamentals remain identical. Even text and dialogue generation already existed in chatbots - I’m thinking particularly of Dialector, programmed by filmmaker Chris Marker in 1988, which already demonstrated a rudimentary form of dialogical reasoning.

The arrival of ChatGPT certainly marks a threshold, an acceleration. But this acceleration comes less from the technology itself than from the sociological event it represents: an experimental system found itself massively and rapidly appropriated by the general public. ChatGPT, despite its initially average results, offered creativity, a capacity for connection-making, and a singular mode of machine reasoning linked to the concept of deep learning. This capacity had already struck minds in 2016 with AlphaGo, which beat humans at the game of Go thanks to its own reasoning modality, a feat much more impressive than Deep Blue’s victory at chess in 1997, obtained by brute force in a much more limited space of possibilities.

Attachment to machines is a reality

The dimension of potential replacement of human relationships by human-machine relationships is not new. Sony’s AIBO robot dog, in the late 1990s, already aroused genuine attachments. Owners mourned the “death” of their mechanical companion when it definitively broke down. More recently, during the transition from ChatGPT-4 to ChatGPT-5, users testified to their dismay at the disappearance of what they perceived as a particular personality, a unique thinking companion replaced by another being.

This capacity for attachment reveals something fundamental about our relationship with intelligent machines. As Sherry Turkle observes in Alone Together: Why We Expect More from Technology and Less from Each Other (2011), we project onto these systems relational qualities that they simulate but do not truly possess. Yet this projection is not necessarily negative if we understand it for what it is: an extension of our human capacity to create meaning and connection, even with non-human entities.

Critics, whom I would qualify as enlightened but obscured by their preconceptions, claim that artificial intelligences reduce our cognitive capacities. According to them, reading AI-generated summaries deprives us of the deep learning that comes from complete reading. This vision seems to me to miss the essence of what these tools can bring us. Let’s delve into the subject.

Freeing time for what’s essential: human relationships

I argue that artificial intelligence, by saving us time and energy on technical tasks like document synthesis, frees us for more relational activities. My intuition is based on more than forty years of international research, relayed in France by INSERM and Santé publique France, which have demonstrated the effectiveness of interventions centered on psychosocial skills for mental health, reduction of risk behaviors, violent behaviors and addictions, as well as educational success. Yet those long hours of work we spent alone, isolated facing heaps of texts, could generate discouragement, feelings of inadequacy or abandonment, particularly among those whom these repetitive tasks deeply bored. How many brilliant people have abandoned their studies or their profession faced with these technical requirements that are cognitive, certainly, but fundamentally repetitive? What a loss for themselves and for the community!

If machines can perform these syntheses and reasoning, which fall under cognitive techniques, our role evolves toward correction, verification, control and adaptation to our specific needs. We explicitly gain time, time we can invest in richer human interactions. This time saved on unstimulating cognitive tasks becomes time available for connection, offering us more mental space and less stress, thanks to these augmentations of our capacities by the machine.

Allow me to share a concrete experience. I recently accompanied a group of high school students in organizing an itinerant screening of short films they were making themselves on the theme of ecology. This screening, included in a film festival program, had to be held in public space with a real audience and real stakes. Time was limited between the start of the school year and the festival a month later. The young people had to not only research, find images, do their editing, write their texts, but also imagine the theatricalization of their screenings: what to say before, what to say after, how to coordinate those handling technology, sound, projection, audience support. All this required time, commitment, attention and self-confidence - deeply human and relational skills.

AI as a factor of inclusion and confidence

I advised them to use artificial intelligences without hesitation for their research and documentary syntheses. This allowed them not only to save precious time but also to build better-informed and more interesting content for their spectators. The facilitated access to information enriched their presentation without exhausting them in preparatory tasks, allowing them to focus on the essential: the show offered to the public, which requires self-confidence and time to journey within oneself to become capable of daring to express oneself and stage oneself.

The question of self-confidence deserves particular attention. People who struggle to write or make many spelling mistakes gain enormously in their social relationships by being helped by artificial intelligences. Contrary to received ideas, this does not divert them from developing their capacities. Quite the contrary: freed from the stigmatization linked to their language difficulties (for which the deplorable quality of the educational system is more responsible than they themselves), they access a new social legitimacy.

This legitimation allows them to authorize themselves to implement their true capacities, particularly in writing. Spelling difficulties often stem less from real lack of abilities than from internal blocks, emotional brakes inscribed by years of stigmatization and judgments. In terms of psychosocial skills, AI helps these people reduce what researchers call “experiential avoidance,” this tendency to flee psychological difficulty rather than confront it. Thanks to AI support, they can finally welcome their internal experience without fear of judgment, thus creating the conditions for true psychological safety necessary for all learning.

As Pierre Bourdieu showed in Language and Symbolic Power (1982), mastery of legitimate language constitutes symbolic capital that largely determines our place in social space. AIs can democratize access to this capital, thus participating in the objective of emancipation and empowerment that is at the heart of psychosocial skills development.

Toward an augmentation of psychosocial skills

People who get help from artificial intelligence therefore gain in skills thanks to this new legitimation. We touch here on psychosocial skills (PSS), absolutely central to what makes our humanity. These skills, defined as a set of cognitive, emotional and social knowledge and know-how, allow each individual to maintain a state of psychological well-being and promote their mental health from a global perspective. Our educational system has for too long concentrated on academic skills - mathematics, French, physics, etc. - to the detriment of these fundamental relational skills.

Psychosocial skills constitute the foundation on which we build humanity. They are articulated around four major functions, called the “Subject Self”: Attention, Analysis, Decision and Action (AADA), mobilized in all our conscious and voluntary actions. They include self-knowledge, critical thinking, understanding and identification of emotions, effective communication, the ability to develop positive relationships and act prosocially. Even more deeply, they involve the ability to welcome one’s internal experience, accept one’s thoughts and emotions, and not flee psychological difficulty but confront and regulate it.

Artificial intelligences, by freeing us from the weight of certain technical tasks, paradoxically become factors for augmenting these psychosocial skills. They allow us to devote more time and energy to activities that develop these skills: daily interactions valuing kindness, gratitude and mutual aid, reformulation and active listening exercises, sharing feelings. In the case of the high school students I was accompanying, AI allowed them precisely to concentrate on these relational dimensions - coordinating their actions, expressing their creativity, developing their self-confidence - rather than exhausting themselves in documentary research.

This perspective joins Douglas Engelbart’s vision on augmenting human intellect (Augmenting Human Intellect: A conceptual framework, 1962), but also the historical evolution of the very notion of PSS. From the Ottawa Charter (1986), which already evoked health promotion as a process conferring on everyone greater control over their well-being, to the current French strategy aiming for “generation 2037” to be the first to grow up in a continuous environment of PSS support, we are witnessing growing recognition of these skills. Generative AIs fit into this lineage of augmentation tools, not replacement, in the service of individual emancipation and empowerment.

Rethinking our relationship with technology

Our relationship with artificial intelligences ultimately reveals our relationship with ourselves and what makes our humanity. Rather than fearing a fantasized replacement, we could embrace these tools for what they are: means to free ourselves for what is essential. Time saved on technical tasks becomes time invested in human relationships. Energy saved on tedious syntheses transforms into availability for listening and exchange.

Artificial intelligences do not threaten our humanity; they invite us to redefine and cultivate it where it truly expresses itself: in our unique capacity to create connection, meaning and affect. It is in this reconfiguration of our priorities, in this refocusing on the relational, that perhaps lies the deepest promise of these technologies.