Let me tell you a story.  In 1981, a man named Jean-Michel Jarre saw the future of China — machines, music, and myths colliding.  That was the year I was born.  Now, in 2024, I’ve created The Red Dragon, a song and a music video pulled from the depths of the unknown.  The Dragon isn’t just a creature.  It’s AI, rising fast and strange.  It’s China, its cities glowing like electric dreams, leaving the West in the dust.  My video?  Built with AI tools — beautiful, broken, and bizarre.  Like directing a dream, but your dream doesn’t always listen.

What came out of it?  A four-minute visual journey that’s surreal, shimmering, and just a little unsettling.

A case study too — for those curious about the weird world of creating art with machines.

In early January of 2025, I embarked on this project using the most advanced AI tools available to the public at the time. After a week of testing various models, I identified the ones worth using and those to avoid.  

Current AI models fall short of producing clear, logical animations, such as explainer videos that demonstrate simple concepts (e.g., layers of a mattress deforming under weight or bacteria being cleaned off teeth). Instead, AI excels in creating rich, detailed, and surreal worlds, akin to a digital hallucination inspired by Hieronymus Bosch.  

Due to this, my goal was to craft a dreamlike, allegorical  video rather than force precision from the tools. Even with premium features, the process often involved long delays, frequent downtimes, and unpredictable results. For a five-second animation, rendering could take 10-30 minutes, with no guarantee of usability. To stay productive, I worked on multiple scenes simultaneously, vetting animations and prompts while waiting for others to render. This chaotic workflow required meticulous organization, with folders filled with imperfect results that I hoped to salvage during editing.

Since the dawn of civilization, humans have pursued the elusive state of flow—a profound sense of immersion and connection. This universal longing has manifested through prayer, meditation, the use of plants, dance, music, and more recently, technology. Each method serves as a portal, offering a temporary escape from the mundane and a glimpse into transcendence.

In the digital age, the pursuit of flow has taken a new form: the digital flow state, a dopamine-fueled rush that captivates and consumes.In my music video for The Red Dragon, this concept is brought to life. It depicts a world where transhumans — beings fused with technology—are drawn into a collective levitation, magnetized by the promise of an apparent blissful state. Is this a genuine state of transcendence, or merely a technological imitation of true bliss? The ambiguity invites the viewer to question the nature of the experiences we seek in a hyper-connected world.

As the video unfolds, it delves deeper into the tension between liberation and dependency. The collective levitation symbolizes the overwhelming pull of technology, uniting individuals yet erasing their distinctiveness. The vivid, hypnotic imagery serves both as an homage to humanity’s unending quest for transcendence and a cautionary exploration of its potential cost.

Using text prompts as the primary interface for directing animations feels limiting. While it simplifies the process for beginners, it sacrifices control. The experience can be likened to directing a film crew via text messages from thousands of miles away. Modifications are nearly impossible; any change generates an entirely new representation, making iterative refinements frustratingly unattainable.

Traditional video production—with real cameras, actors, and locations—offers more control but demands significant time and resources. In contrast, AI provides visually stunning results with minimal effort, but its lack of flexibility and precision can be frustrating. For this project, I didn’t have the budget or time for a traditional approach.

Given the nature of streaming platforms, where content often feels disposable, AI served as a practical surrogate. In general, on cellphone’s small screens, AI-generated visuals can rival  with traditional footage.

The song production was a project and a process in itself. My main DAW for music is Reason Studio, which resembles modular synthesizers in its architecture, interface, and usability. This setup allows me to quickly sketch ideas—short loops or bars with a catchy sonic quality. I save these ideas and revisit them later with fresh ears, deciding which ones still resonate with me. If a loop triggers something strong, I begin building the full song around it.

Writing electronic music is a process intertwined with sound engineering. Unlike a guitar soloist who lives in the moment electronic music production requires juggling constraints, inputs, outputs, and connections. The final song is an intricate web of interconnected instruments and channels, with numerous knobs and parameters keyframed—like a massive After Effects project.

To create the four-minute video, I relied on After Effects to address AI-related issues: smoothing transitions, hiding glitches, and time-warping animations. It felt akin to working with self-generated stock footage—tailored to my vision but constrained by flow and angles. The result wasn’t perfect, but it captured the dreamlike essence I sought.

Creating a music video with AI in 2024 is an exercise in balancing innovation with compromise. While AI offers unparalleled creative possibilities, it comes with significant limitations. For artists with constrained resources, it provides a viable path to produce visually compelling content. However, the lack of control and iterative capabilities remains a challenge.