I'm in Cannes for a few days as part of my other activity, the new film production studio The Media Company, which I co-founded with Didier Lupfer and Édouard Boccon-Gibod. I'll say more about it soon if you're interested.

As you know, cinema has entered an incredibly complex transformation. On one side, the economic crisis is forcing nearly everyone to turn to more conventional films, with tight budgets. And at the same time, because of a moral and intellectual crisis, the biggest hits (namely Adolescence, Severance, co-written by my friend, the very talented Wei-Ning Yu, and of course Andor) have shown that audiences want, and reward, extremely demanding and complex series.

In an increasingly complex world (and that, incidentally, is why I launched this newsletter), the last thing people want is to settle for simple things. This isn't a rhetorical question but a simple anxious reflex: we force ourselves to rationalize what is happening, sometimes before digging into it.

In the past, the press and television played that role, notably through their columnists. Whether you agreed with them or not, they knew how to point to what mattered, which forced us to think. Today, television serves many functions, but it is no longer really a space for that.

What has largely replaced political criticism today is comedians and artists, especially in the United States. Stand-up acts that are, to be honest, the only ones really taking it on. But there is also cinema. And what is fascinating is that cinema, and series in particular, has never been so political. Who would have thought that Andor, a spin-off series from the Star Wars universe, would evolve toward such political density? It shows the making of a monster, the Empire, through characters completely overwhelmed by the death machine they help to run in spite of themselves. This project, let's remember, is produced by Disney. Disney which, incidentally, is now panicking, because the success of Andor is so massive that it has no intelligent series in reserve to take over from it.

The equivalent of Andor for cinema is obviously The Brutalist. I had the chance, yesterday, to talk with one of its co-producers. What fascinates me, and what I pointed out during our conversation, is that sometimes a single bold move can reconfigure an entire sector. The success of a film like The Brutalist goes far beyond that of an isolated work: it is the triumph of a courageous editorial choice, in a landscape that is increasingly standardized, even sanitized. It is a statement that a film capable of defying the codes, the formats and the industrial logic can not only exist but also find its audience. And from there, it becomes much more than an artistic object: it acts as a revealer, a force that compels the whole system to reposition itself, to rethink its standards, its narratives and its models.

Barely arrived in Cannes, I saw how much cinema is facing new constraints, but also a genuine opportunity for transformation.

In the world of cinema and artificial intelligence, which is in the spotlight here as you can imagine, two major approaches coexist.

The first is to create a new cinema, built entirely on AI.

Mathieu Lorrain, with whom I had the chance to talk (someone I like and watch with great interest), is the creative face of DeepMind. What he explains is fascinating: the technology changes, but the fundamentals remain. He says that story remains the constant in a world in full disruption. Beautiful stories, good stories.

One of the questions I asked during the panel, and one that affects us all, is this: why and how, in a world where you can re-prompt (that is, ask ChatGPT-3.0 to recover the original prompts used by artists, then reproduce their works identically, or make them evolve), does the quality of what we see risk stagnating? It will become very hard to innovate, to be truly different, in a world where everything becomes standardized and copied.

What is certain is that the advertising world has already seized on these tools. And a thirty-second commercial produced by AI, with a few minor visual concessions, is already a reality.

A word of caution: making a film with AI takes an enormous amount of time. Kevin Abosch, one of the pioneers of the field, was in Cannes with his film Shitcoin, which unfortunately I wasn't able to see. He explained to me in the past that he builds his own models, and that, fundamentally, the key to everything being done in this field lies in creating custom models.

My feeling is that AI today, measured in tokens, is extremely expensive. And if you measure it by the monthly cost billed to the user, you realize that Google, Runway and the others have to invest massively because everything is running at a loss. On one side, there are the semi-professional users, who experiment with these tools within a constrained creative framework and will use up their credits, but rarely go beyond, for lack of budget. On the other, the professionals, whom the platforms hope to win over with framework agreements spanning several films, betting on their capacity to spend fortunes in tokens to industrialize production.

The second approach, the one we rely on, is to integrate AI at the heart of the filmmaking process, starting in preproduction.

In a world under financial and time constraints, this makes it possible to do incredible things.

For the past six months, we have put in place, within the studio, a completely new way of working. It is largely inspired by what I do in my research for this newsletter, which I write alone, but with the help of some fifteen intelligent agents (a sort of Central Intelligent Agentics) that crawl the web around the clock.

For our films, AI lets us take the creative proposition all the way. We can imagine previsualizations, scenes and, above all, genuine collaborative work from the very start between the director, the screenwriter, the film crew and the artificial intelligence. Using AI to co-brainstorm is, in my view, fascinating, and still largely underexploited.

Someone summed it up very well: in the world of AI, either it is used as a tool, or it is treated as a collaborator. For my part, it's the second approach that interests me. A collaboration with remarkable creative intelligences, where AI is allowed to propose ideas.

So yes, the question today is learning to master these tools. But to come back to what I was saying earlier: using them only to extend narratives or worldviews that are already standardized would be to miss their real potential.

What I expect from AI in cinema is not that it disrupt for the sake of disrupting, or shake up the market for the pleasure of shaking it up. What I expect is that a film made with it, fully and lucidly, will be the one to change the rules of the game. As The Brutalist did. As Andor did. Like all those projects born of misunderstood bets, that nobody was expecting, and that, once released, immediately join the pantheon of the best films.

For the next few days, I'll keep being woken up by the sound of the waves in the morning. An experience I admit is very pleasant. If you're a member and you're in Cannes, a meetup is obviously in order. Otherwise, see you very soon for the next newsletter.

Become a paying member

What is free no longer has any value. Most of the analysis you read today is biased, filtered by political or commercial interests. By subscribing, you get access to our real-time news feed and our strategic events, and you join the Digital Resilience Council (Conseil de la Résilience Numérique): a space for learning to live in a digital world on the edge of chaos, reserved for those who want to understand, anticipate and act.

To subscribe, click here.