Project Overview

Open Studio: Blending UX and Generative AI for Collaborative Storytelling
 

Open Studio was a live, co-created web comic centered on a futuristic racing narrative. Set in dynamic cities like Panama City, it invited participants to help shape the story by submitting AI-generated vehicles and characters. As the lead designer and strategist, I used this project to explore how generative AI can enhance storytelling, accelerate design, and foster public participation.

I used new AI tools to create images and videos for public-facing content, balancing speed with ethical design choices. I led the full design process—from early ideas to final visuals—using a custom AI workflow to build interactive, engaging experiences.

This project was created in 2023, before the most popular genAI tools had more consistent results. 

Role

AI Trainer
Art Director
QA
Social Media
UX Designer
UX Writing
Video Editing

Team

CEO/Copywriter
Developers

Deliverables

Experience Design
GenAI Artwork
Social Media
Visual mocks
Wireframes

Problem statement

Can we convince people to use genAI tools to create a collaborative web comic?

This project faced some challenges:

• Inconsistent Visual Outputs from AI Tools
  Struggled to generate a consistent character appearance, especially for a biracial identity, even with fine-tuning.
  
• Inconsistent Visual Style from User Submissions
  Users submitted images using different AI tools, which led to clashing styles between characters and cars. To keep the visuals more consistent, we encouraged everyone to use our custom image generator.
• Limited Ethical AI Tools at the Time
  For the public-participation part, we wanted to avoid using AI tools that allowed users to copy artist styles, which narrowed visual options but upheld ethical standards. For the part I created, I used prompts that were in a certain style of games but avoided copying individual studios' or artists' styles.
• Generative Video Tools Were Not Production-Ready
  AI video platforms were unreliable, so I relied on traditional editing tools for timely and polished outputs.
• Explaining a New Concept to New Audiences
  Had to clearly communicate a complex idea (AI + web comic co-creation) to users unfamiliar with both.

It took several rounds to get alignment on the main character’s design across the team and with the CEO. Kenji, the protagonist, is mixed race. As someone who is also biracial, I saw early on that the AI models had trouble representing that identity accurately.

I used more advanced prompting to guide the outputs toward a more accurate and respectful portrayal, while working around the model’s built-in limitations.

Visual Design

Designing a Friendly, Futuristic UX

To attract users unfamiliar with AI or web comics, I designed a marketing site that felt inviting, playful, and futuristic.

• Content Design: Used ChatGPT to co-write clear, inclusive site copy that explains a complex idea in simple terms.
• Visual Direction: Created moodboards and style references in Midjourney to visualize the look and feel.
• Tool Constraints: Since tools like Adobe Firefly weren’t fully ethical or available yet, I chose image models and prompts that didn’t mimic specific artist styles.

I drafted the initial copy using ChatGPT, then refined it with input from the CEO—who was also writing the main storyline—and the rest of the team. Collaborating with multiple voices led to some sections becoming a bit wordy by launch.

Visual Experiments

Cars, Characters, and Cohesion

To ensure stylistic cohesion between user-submitted cars and AI-generated avatars:

I tested multiple approaches, including:

• Blending character and background images using Midjourney prompts
• Generating visuals in composite workflows using Photoshop and Leonardo
• These tests helped map out the limits of visual consistency with current-gen AI tools.

I tested ways to blend illustrated and photorealistic styles. Since the focus was on visual style, I didn’t worry about perfect accuracy in the renderings for this experiment.

Social Media

Creating Assets for Social Storytelling

I produced short videos and graphics to promote the comic and highlight submissions.

Challenge: AI video tools were slow and inconsistent.
Solution: Used AI for image morphing (Hugging Face) but stuck with traditional tools like After Effects and Figma for video edits.
This hybrid workflow let us iterate fast while keeping the quality and tone consistent.

Some of the social media banners I created for the races.

AI Workflows

Impact

• Faster Turnaround: AI tools cut our asset creation time by 50% while supporting creative experimentation.
• Inclusive Design: Prioritized ethical image sourcing and visual equity, even with emerging tools.
• Scalable Pipelines: Built reusable AI workflows that can expand to new storylines or formats.

AI Workflows

Why It Matters

Open Studio wasn’t just about fun storytelling—it was a real-world testbed for how AI can supercharge creative pipelines, align with ethical design, and make content co-creation more inclusive.

This project shows my ability to:

• Prototype, test, and scale generative design tools
  Bridge UX, AI, and visual storytelling
• Adapt workflows to evolving platforms and public engagement needs
• Perfect for teams exploring how AI can power the next generation of creative tools.

Similar Projects

© 2023 Danita Delce. All rights reserved.