ASHEN
Ashen - Kinetic Ghosts
๐ฅ AI-powered music and video
Ashen is a fully fictional, AI-generated alt-rock persona created for artistic experimentation. All songs, lyrics, visuals, and lore were generated using AI tools. This page showcases the music, the character, and the creative world built around Ashen.
About This Project
Ashen is a fully fictional, AI-generated alt-rock persona created for artistic experimentation. All songs, lyrics, visuals, and lore were generated using AI tools. This page showcases the music, the character, and the creative world built around Ashen.
Artist Bio
Ashen is a Canadian alt-rock vocalist forged in smoke, steel, and heartbreak. Her music blends blues grit, modern rock intensity, and cinematic emotional storytelling. With a smoky, powerful voice and lyrics that cut like broken glass, Ashen embodies motion, memory, and the emotional physics of heartbreak.
Her debut fictional album, Kinetic Ghosts, explores themes of wildfire, asphalt, phantom weight, and the forces that pull us apart or hold us together. Though Ashen is not a real artist, her world feels hauntingly authentic โ a universe built from embers, neon reflections, and the ghosts of what once was.
Songs & Lyrics
Embers on the Asphalt
Fingers of Lead
Amber Eating the Sky
How Ashen Was Created โ Music, Lore, and AI Video Production
Building a Fictional Rock Artist With Modern AI Tools
Ashen isnโt a real singer โ but she feels like one. Her songs, her voice, her world, and even her music videos were created using a combination of AI tools, creative prompting, and a surprisingly simple workflow. This project started as an experiment: Could a fully fictional altโrock artist be built from scratch using accessible AI tools? The answer turned out to be yes โ and the process was far more fun than complicated.
This article explains exactly how Ashenโs music and videos were created, stepโbyโstep, using tools anyone can access.
๐ค Step 1 โ Generating the Music and Lyrics
The foundation of Ashen began with a single prompt:
โA song about heartbreak in the style of Ruby Black. Driving rhythm and powerful lyrics.โ
Using Gemini (Lyra 3), that simple prompt produced the first full song: โEmbers on the Asphalt.โ
The result was emotional, bluesโrock inspired, and lyrically cohesive. That success led to a second step: asking an LLM to help brainstorm a full album concept. The model suggested themes, track ideas, and stylistic directions that matched the tone of the first song.
From there, additional songs were generated using iterative prompts, refining the emotional tone and imagery until Ashenโs voice felt consistent across tracks. The three initial songs:
- Embers on the Asphalt
- Fingers of Lead
- Amber Eating the Sky
formed the core of her debut fictional album, Kinetic Ghosts. Each song was created with minimal prompting โ the key was consistency in tone, imagery, and emotional direction.
๐ถ Step 2 โ Creating Ashenโs Persona and Lore
Once the music existed, the next step was to build the artist behind it. Using an LLM, Ashenโs persona was crafted: a Canadian altโrock vocalist forged in smoke, steel, heartbreak, and wildfire imagery. Her lore, aesthetic, and emotional universe were shaped to match the themes in her songs โ asphalt, embers, phantom weight, gravity, and fire.
This persona became the backbone of the project, giving the music a sense of authenticity and narrative cohesion.
๐ฅ Step 3 โ Generating the Music Videos With LTX 2.3
With the songs complete, the next challenge was creating visuals. The videos were generated using LTX 2.3, running locally on a Windows machine with an NVIDIA 3060 GPU. The setup was done through Pinokio and WAN 2GP, which made installation and model management straightforward.
The video workflow:
- Each video was generated in 10โsecond segments, allowing for more control and easier iteration.
- Prompts were crafted to match the lyrics and emotional tone โ embers on wet asphalt, wildfire skies, ghostly hands of smoke.
- Segments were rendered individually, then reviewed for consistency.
- Once all segments were complete, they were stitched together into full videos.
๐ฌ Step 4 โ Editing and Final Assembly
The final step was combining the audio and video. All editing was done using Windows Clipchamp, which made it easy to:
- align the video segments with the music
- add transitions
- adjust timing
- export in high quality
Clipchampโs simplicity made it ideal for assembling AIโgenerated footage without needing professional editing software.
๐ Step 5 โ Publishing and Sharing
Once the videos were complete, they were uploaded to YouTube. This page showcases the songs, lyrics, and videos together โ forming a complete multimedia experience for Ashenโs fictional debut.
The goal isnโt to hide the fact that Ashen is AIโgenerated. Instead, the project embraces it: a creative experiment in storytelling, music, and visual art using modern tools.
๐ก Why This Project Works
Ashen resonates because she feels real โ not because she is real. The combination of emotionally charged lyrics, cohesive visual storytelling, a strong fictional persona, and a transparent creative process creates a unique experience that blends human creativity with AI assistance.
This project shows that anyone can build a fully realized fictional artist using accessible tools, imagination, and a willingness to experiment.
๐ Whatโs Next
More songs. More videos. More lore. A full album. A visual universe. Ashen is just getting started โ and her world will continue to grow as new tools, prompts, and ideas evolve.