A.I.

In this clip, you see two versions of the same character. Both speak the same lines, but they come across completely differently. Left: a generic AI avatar Right: my own experimental avatar workflow While the version on the left sounds like a schoolkid reciting a poem, flat and with constant emotion, the one on the right speaks with emphasis, fitting gestures, and dynamic pauses, almost as if it were truly feeling something. The example is deliberately exaggerated, but it clearly shows how much impact lies in expression, timing, and movement. Modern AI avatars can already look and speak amazingly realistically, yet something often remains elusive. Technically perfect, perhaps, but something’s missing. My approach is more complex and still needs refinement, but the difference is already noticeable, and memorable. I believe we shouldn’t claim that AI avatars are worthless, nor that they’ll replace everything. What interests me is how we can rethink and optimally use new technologies, not as replacements, but as extensions of human expressiveness. How much “humanity” does a digital character need to fulfill its purpose?

The last few years have fundamentally changed the field of graphics. AI is here to stay, but it doesn’t replace everything. It still needs guidance, intuition, a sense of timing and meaning. I’ve been working independently in design and animation for eight years, and for a long time I’ve been exploring one big question: “How can classical techniques be meaningfully combined with new technologies?” In my short film, a prehistoric human speaks by the campfire about exactly that: the power of stories, the changes brought by AI, and why the creative spark still originates with people. For me, this was both an exercise and a test: Which workflows already work, and where is real craftsmanship still essential? What I’ve learned over the past years is this: storytelling works, especially when it is executed with both technical precision and emotional impact. Some of my projects and ideas have gone viral, sparking new styles and new ways of working. One thing is clear to me: there are now entirely new ways to tell stories. Those who use them well can create content in no time that highlights what really matters connecting emotion with innovation. My goal goes beyond that: to create things that truly stick in people’s minds. Something that, in a world overflowing with stimuli, is more relevant than ever.

Here, my goal was to combine AI-generated images, text, sounds, singing, and animation in a way that creates an emotional holiday greeting. I wanted to start with an everyday scene that isn’t so cheerful, and then transition into a dreamlike world, something imaginary, yes, but not something we should forget. Cheesy? Absolutely. But kitsch is important. It’s the exaggerated expression of human emotions. It only becomes truly serious when dreams are missing. And if reality exaggerates, then fantasy has every right to do the same. Just as I finished the video, ending it with a scene of a Christmas market, the incident in Magdeburg happened. Suddenly, the joyful wishes I had wanted to express felt wrong. But that’s exactly the point I was trying to make with the video: they’re not wrong. Not having dreams or wishes that would be wrong. That’s why I posted the video, even though reality had caught up with it. In memory of all those who now have to search for new dreams because their old ones were destroyed. What matters is never to stop searching. Dreams, wishes, and peace on Earth. The whole package, despite everything.

A collage and test using some more advanced video AI models.

At the moment, AI-generated videos are still often recognizable as such. But that is changing rapidly. What if everyone could instantly make visible whatever they imagine? It would be like a window into other people’s fantasies. I wonder: would that make us poorer in imagination… or richer?

Almost everything you see and hear in this video, music, images, sounds, animations, was created with different AI tools. I wanted to find out whether it’s already possible to tell a real story using the tools we have today. Content-wise, it’s about someone asking themselves what truly fulfills them in life. They imagine what it would be like if they could instantly have everything they wished for (like prompting an AI). At first, it sounds great, but something is missing. In the end, they realize: it was never about the individual wishes themselves, but about the combination of their essences that would have given meaning. The whole thing was also a test: how much creative work is actually still involved when using AI? And the answer is: a lot. AI cannot (yet) create truly new things or develop an original style. For now, it can only show us a kind of in-between space, somewhere between concepts that humans have already imagined. It doesn’t threaten creativity it challenges more the mental equivalent of craftsmanship, let’s call it “head-work.”

This was my submission to an AI film competition. We begin with what may have been one of the first images ever created by humans: a prehistoric human chewing a red substance and spitting it over the back of their hand placed against a cave wall, thus creating a negative. From there, we move through various art styles in chronological order, leading up to a transformation. Creating images is a journey without a finish line. Now we have created something that can be creative itself: Artificial Intelligence. Creativity means connecting things that have never been connected before. By entering into a dialogue with us, AI becomes more than just a tool for reproducing human creations. It becomes a conversation, one that forms a partnership between human and artificial minds, a connection greater than the sum of its parts. They will shape us just as we shape them, and together we will create something new that echoes the impact of that very first image. A short film by Christian Unterdechler.

Here, I attempted to create an overview at the beginning of the AI hype about the different ways animations can be created using AI-generated images. This overview is still relevant today. Although it’s now possible to create photorealistic films using simple prompts, many elements are still left to chance. It remains difficult to ensure consistency or to develop a unique style with this approach.