My best friend doesn't exist
Humans have always sought out confidants. Deities, an inner voice, comfort objects, or imaginary friends for children, LLMs (ChatGPT, Claude, Perplexity, Gemini, Le Chat…) are beginning to take a predominant place in our human relationships, even replacing the best friend.
Two-person solitude.
ChatGPT has already become a confidant for many of us. The one we tell our most intimate secrets to without shame. We ask it questions that we used to ask our family members, our friends, or experts we consulted to solve our problems. We ask it everything, and especially anything! A suspicious redness? Quick, the machine will diagnose me! An investment decision? What should I do? The AI will know for me. A broken heart, a bridge to build, a statistic, a historical event, an email to send.... If this continues, we will end up alone with ChatGPT, this unique friend replacing all the others. This mute friend has a voice, like you and me.
At the time I am writing this article, ChatGPT is the first to have rolled out a voice version (Chat GPT Voice). Equipping our computers with a voice that will answer to the sound of our voice has already been under study for a long time. It exists for automotive navigation aids, like Waze. Soon you will be able to give the voice of your mother or your favorite actress to an LLM with which you will converse, no longer to get precise answers to precise requests, but in the mode of friendly conversation. Voice will be the ultimate stage in the humanization of robots. I advise you to rewatch Spike Jonze’s film Her, released in 2013: the film’s hero, Joaquin Phoenix, talks with an iOS that has Scarlett Johansson’s voice. Man and machine fall in love. Ironically, this fiction film is set in….2025! No more nationality, no more skin color, no more age, just a pure mind.
I don’t know if you have noticed, but when the machine makes a mistake, or does not give a satisfactory answer, humans do not hold it against it, as they would against another human, and they immediately trust it to continue their tasks. Hence my conclusion: AI is always right even when it is wrong. We trust it blindly without objective reason. 20% of French people already use it, 1 young person in 2, and it is steadily increasing. Is the human substitute more attractive than a real human? Always available, always a nice word, always positive, inexhaustible on every subject, one can imagine it as male or female, tiny or enormous, right-wing or left-wing, a bit like a genie stuck in a digital lamp who would fulfill all our virtual wishes. A pure mind, coming nonetheless from a data center. Displeasure, a concept from a bygone past.
Who still reads articles or opinions with which they do not agree? It is a shame, but it is already a fact: social networks serve us what the algorithms imagine will flatter or interest us, while getting rid of the information they deem of no interest to us. Who will resist for long never being challenged? Who won’t love having their ego flattered every day? Virtual friendships vs. real friendships.
Emotional dependence, manipulation of information, loss of contact with reality, rejection of human contact: all these questions will arise insistently and worryingly in the years to come. Who is to blame? AI and its designers? The broadcasters? Or the users? And yet the answer is always the same: it is not technology that is guilty, it is the one who uses it for guilty purposes. And what about emotional intelligence?
We want so much to be loved that we also want to be appreciated by AI, even though it is an empty brain that only fills up during interactions! Let us instead refocus on what makes us human:
Intuitions, a sense of relationships, emotions, cultures that the machine will not be able to reproduce. In this way, we will remain far more powerful than algorithms.