From Broken Arrow to Warsaw Glitch: a filmmaker between craft and generation

From television and documentary filmmaking to generative cinema, Michal looks back on the creation of Warsaw Glitch, Broken Arrow, and the current limitations of the AI market.

You come from television, journalism, and documentary filmmaking, but also from video games, since you started at Sony Computer. What has that journey taught you about the way you look at reality, now that you also create fictional worlds?

During my first year at university, we studied ancient philosophy. I don’t know whether you remember Plato’s Allegory of the Cave.

We all live inside a cave, watching shadows projected onto a wall. Then one day, we free ourselves from our chains, discover the outside world, and realize that an entirely different reality exists, even though the only thing that has changed is our perspective.

Then we try to bring that perspective back to the others, but they still cannot see the sunlight. They continue watching the shadows.

I have always thought about that, even though, to be honest, I have never fully understood it.

With AI, I feel as though we all live inside a kind of intersubjective matrix. Not in the sense of, “Dude, we live in The Matrix!” but in the sense that I believe we partly create our own reality.

And with AI, we can now see very clearly that we are capable of creating entire worlds from a small room, sitting behind a computer.

At first, there was curiosity and the desire to transform what you had in your head. When did you realize it was no longer a game, but an actual means of production?

I think it happened when I needed an establishing shot for a documentary.

I thought, “My God, I absolutely need this shot.” At the time, we were using the early versions of Haiper or Luma. I wondered whether I could simply generate a fairly static shot with just a little movement in the leaves or the environment.

I tried it and... it sort of worked.

I thought, “Okay, I can put this exact shot into my documentary. It may not be perfect, but it’s a documentary... maybe no one will notice.”

That was when I understood that it could genuinely become useful within a production process.

Let’s talk about Glitch. You are leading the project, but you obviously have a team around you. How did this whole venture come together?

I built the website myself, but I have some truly wonderful people helping me.

A lot of other people abandoned me, to be honest, because they did not believe we could do it. Some canceled their trips, others pulled out of the project. They told me, “Dude, you are completely insane. This probably looks like a scam.”

I traveled to thirty or forty festivals around the world.

When I went to AI on the Lot, I thought, “Look at what they’ve built.”

What impressed me was not only the AI. It was seeing it at the center of a community that was taking shape around it.

At many other festivals, everything is very small. You show a few films, have a beer, and go home.

Everything happens inside a closed ecosystem.

I wanted something else.

From our first conversations, you emphasized the importance of getting different communities talking to one another. There is also a clear intention to build bridges between traditional cinema and new tools.

I think part of the old guard is gatekeeping cinema.

And because newcomers feel excluded, they sometimes end up doing exactly the same thing with their own knowledge and tools.

For me, the future is hybrid.

But I am not simply talking about mixing live-action and generated images. That is just hybrid production.

What interests me is bringing in masters of cinema.

On the Polish side, for example, I have directors who are around 70 years old and have received Oscar nominations or European awards.

I want to put them in the same room as an incredibly talented AI artist in their twenties who does not yet have their expertise. The first group has the craft, but not necessarily the new tools. The second has the tools, but not yet the full craft. I do not want young people to replace the old guard. I want them to mix.

And on top of that, I invite engineers who can explain what is happening to both sides. Many AI creators use tools without really thinking about how they are made, while the old guard has absolutely no idea how they work.

Finally, since we are bringing in engineers and platforms, we are also bringing in investors, CEOs, and founders. We are gathering the entire production chain, from infrastructure to the final product and its distribution.

You have technology partners, engineers, and highly technical guests. The risk is that these discussions quickly become inaccessible. How do you prevent technology from taking control of the event?

The answer is extremely simple: alcohol.

No, seriously, everyone lives inside their own bubble.

I want all those bubbles to melt and mix together. That is why I organize community mixers. There will be a Machine Cinema mixer, for example, with engineers and investors invited. There will also be an investor mixer where we invite members of the creative community.

Each community will have its own spaces, but people from other worlds will be inside them. This kind of cooperation, stretching all the way from computing and infrastructure to the final product, has never, as far as I know, been organized this way in Europe.

And Warsaw, at the heart of Europe, feels like an excellent place to begin this conversation about the future of European entertainment and cinema.

Glitch therefore becomes both a showcase and a marketplace. People can look for work, meet investors, speak with platforms, or connect directly with developers.

Exactly. I have been to Cannes twice.

You have the Cannes Film Festival, centered on creators and films. You have Cannes Next, focused on technology and emerging practices. And then you have the Marché du Film. This year in Cannes, people working with AI were much more visible.

The previous year, they were difficult to find. This year they were there, but they were still separated. I simply thought: let’s put everyone in the same place, lock them inside the same building, and make it easier for them to meet.

What would success look like to you on August 30?

If I manage to deliver 50% of everything I have