🔴 Is AI a Magic Cure for the Mediocrity of the Real?

Artificial intelligence is gradually becoming the central subject of this newsletter, from a broader perspective.

I am preparing my new talk, The Sovereign Way: New Rules for a New Internet, for the fall, which brings in the concepts of cognitive autonomy and post-sovereignty. If you are interested in a presentation, get in touch.

For this early-August newsletter, I offer you a reflection that grew out of a debate about writing with AI. I wish you a good vacation, and we will meet again at the end of August for new newsletters and a few surprises for paying subscribers.

One of the great debates I regularly have with my friends concerns the question of writing and AI. Is it a tool that helps you write, or a tool for editing and finishing?

Starting from a blank page has never scared me, but editing a text so that it can be published has always been a personal bottleneck.

At the opposite end, one of my friends, who has put out many widely noticed articles on LinkedIn, told me he was the exact reverse: ask the machine to write everything, a kind of childbirth, an inverted maieutics in which writing resembles sculpture, where mistakes and backtracking are possible.

What is true for writing is also true for research.

For some people, like me, discovering and annotating documents during research is part of the pleasure. An instant answer never brings the same sense of “feeling the work done,” even if I understand why many people need it.

This question is in its very early days.

It is obvious that this need to sidestep the difficulties of certain kinds of cognitive work is going to become the norm, even if the result does not always live up to our expectations.

When thinking becomes outsourceable

One of the most underrated contributions of generative AI is not that it has transformed our ability to write, but that it has restored a form of confidence to those who did not necessarily have the knowledge or the desire to write.

Hiring managers, unable to tell perfectly written résumés apart, are grateful!

Faced with this tsunami, everyone wants to jump on the bandwagon. AI seduces the consultant running behind schedule as much as it does political leaders, who hope to use these new cognitive capabilities to govern. As if a magic button instantly handed them the understanding of, and the answers to, complex problems.

This desire is not new. It follows directly from industrial offshoring, when part of our elites saw in the closing of factories a form of cognitive liberation, convinced they would no longer have to manage people, workers, unions and their everyday problems.

Alas, that offshoring also made them completely replaceable. If managing people is no longer part of their job, then any manager can handle the abstract management of an industrial system.

What is interesting is that the generation that accelerated the deindustrialization and the de-computerization of France had not understood that most of the value lies in the necessary friction with workers and in the constant improvement of production lines. Tesla, Apple and Nvidia draw their value from industrial processes of staggering complexity.

Whatever one may think of an Elon Musk or a Tim Cook, they have never forgotten that know-how is a daily battle and a constant collaboration between blue collars and white collars.

With the spread of AI, cognitive offshoring also carries the risk of no longer knowing what to do, lost in data and AI analyses too far removed from reality. But also a hope.

This cognitive offshoring is no longer feared if it becomes an escape, if only for a moment, from the mediocrity of the real.

Refusing the mediocrity of the real

“We live in a world where disappointment has become a form of collective experience.” Eva Illouz, Explosive Modernity

Reading Explosive Modernity forces us to analyze and update this phenomenon. This refusal, embodied by Emma Bovary in Flaubert's novel, gets a fresh gust of air from generative AI, which promises everyone an improved version of themselves, of their speech, of their thinking.

Where Emma Bovary dreamed through books, we dream of whatever our screens, social networks, apps and self-optimization interfaces spit back at us.

To the point that for many, daily life becomes unbearable not because it is painful, but because it seems dull, slow, opaque. What Flaubert described as a bourgeois drama has been modernized: real life always seems less desirable than its augmented digital version.

In this context, AI plays an ambiguous role. On one side, it promises to help us “write better,” faster, more clearly, more elegantly. On the other, it reinforces the idea that our raw thinking is insufficient. Behind every generated text there is an implicit negation: we are not enough.

It promises to help us write better, but at the same time it weakens the very idea that our thinking deserves to be heard as it is.

AI thus becomes the contemporary extension of Bovarism: a tool that improves self-expression while heightening the feeling of inadequacy with the real.

It feeds the idea that mediocrity can be circumvented not by accepting it, but by rewriting it. Writing, a slow and therefore inevitably disappointing activity, becomes a frictionless interface for the performance of the self. Just look at our LinkedIn newsfeeds.

But all this also reveals an instability: the “better” we write, the less we recognize ourselves in what is written.

AI does not give us access to ourselves; it pushes us to outsource our subjectivity, to delegate it to a system supposedly better than we are. In a way, ChatGPT has transformed writing the way Instagram transformed our view of reality. A form of staged perfection.

That may be the real stake of this cognitive offshoring: a world where everyone is a “researcher” without researching, a scientist without knowledge, and above all a leader without ever confronting difficulty.

The coming decade will also be a decade of frustration, in which we collide with the asymmetry between our augmented aspirations and a raw, imperfect reality, with the risk that AI does not eliminate mediocrity but merely tries to erase its pain.

Paying subscribers section