Maya and Ghost: Seeds of Tomorrow

Inspiration Our inspiration came from a deep concern for the planet, contrasted with an enduring belief in hope and the power of connection. We wanted to explore a classic post-apocalyptic theme but focus it not on the nihilism, but on the resilience of nature and the potential for redemption. The core idea was to create a powerful, emotional narrative around an unlikely bond—a lone scientist and a majestic tiger—as the last symbols of life fighting back against a world ruined by human greed.

What it does Maya and Ghost: Seeds of Tomorrow is a cinematic proof-of-concept and a narrative short film. It serves as a visual and emotional introduction to a larger story. Its purpose is to leverage AI-driven visual generation to tell a compelling, feature-film-quality story in a condensed format. It showcases how disparate AI tools can be wrangled to create a cohesive narrative that explores deep themes of survival, interspecies trust, and humanity's final chance at redemption.

How we built it This project was a technical and artistic marathon, built using a hybrid workflow that bridged multiple "early-generation" AI models with traditional post-production techniques.

Conceptualization: We started with the core story, outlining the key scenes, emotional beats, and character arcs for Maya and Ghost.

Asset Generation: We used a suite of different AI image and video tools to generate the visual components. This involved leveraging the unique strengths of each platform:

Kling & Hailuo: Utilized for generating specific video sequences and motion.

Gemini & Grok Image: Employed for creating high-fidelity still images, character concepts, and key environmental art.

Editing & Composition: All the generated assets were imported into Adobe Premiere Pro. This is where the real "stitching" happened—we meticulously edited the clips, blended still images with motion, and performed colour grading to create a sense of visual continuity. The soundtrack and sound design were also added at this stage to bring the world to life.

Challenges we ran into The entire process was defined by overcoming the limitations of the AI tools available at the time.

Character Consistency: This was by far the biggest challenge. Maintaining Maya's and Ghost's appearance across different scenes and different AI models required countless iterations and clever editing tricks to hide inconsistencies.

Visual Cohesion: Each AI tool had its own aesthetic. Blending shots from Kling with images from Gemini, for example, required significant work in Premiere Pro to make the lighting, colour palette, and textures feel like they belonged in the same universe.

Fragmented Workflow: Juggling multiple platforms was incredibly inefficient. We had to constantly switch between different tools, each with its own interface and quirks, making the creative process slower and more complex than a single, unified solution.

Accomplishments that we're proud of Despite the challenges, we are incredibly proud of creating a finished piece that is both emotionally resonant and visually compelling. Our main accomplishment was proving that a powerful, coherent narrative could be told with these fragmented tools. We successfully established a tangible sense of a bond between Maya and Ghost and built a believable, atmospheric wasteland—achievements that were a real struggle with the technology of the day.

What we learned This project was a masterclass in creative problem-solving. We learned that the artist's vision is the most crucial element; the tools are secondary. We mastered the art of "AI-wrangling"—knowing which tool to use for which specific task and then using traditional software to force them into a cohesive whole. Most importantly, it gave us a profound appreciation for how rapidly AI technology is evolving, as workflows that took us weeks of painstaking work can now be accomplished much more easily.

What's next for Maya and Ghost: Seeds of Tomorrow The future is incredibly exciting, especially with today's advanced AI tools.

The Definitive Remake: Our immediate goal is to re-create the short film using the current generation of AI video models. This will allow us to achieve perfect character consistency, more dynamic action sequences, and an even higher level of cinematic polish.

Episodic Expansion: We plan to expand the story into a mini-series, with each episode focusing on a different challenge or faction within the wasteland. This would allow us to explore the lore and world-building in much greater detail.

Interactive Story: We're exploring the idea of an interactive narrative or graphic novel where the audience can make choices for Maya, influencing the journey and its outcome.

Built With

• adobe