Discover Your Potential: Insights from a Hollywood Creative Director by Lee Heebok

Hollywood title-sequence veteran Lee Heebok — creative director on Spider-Man 3, Superman Returns, Speed Racer and Ninja Assassin — teaches his design principles and generative-AI workflow for the first time. Across 41 lessons / 19 hours 50 minutes you go from classic design theory (fractal, contrast, rhythm, abstract visual storytelling) to hands-on production in ChatGPT, Midjourney, Artbreeder and Stable Diffusion, finishing with a complete fictional film title-sequence styleframe set built from a virtual scenario.

🎯 What you’ll learn

• Apply design principles — fractal, contrast, rhythm — and abstract visual storytelling (theme, motif, symbol) to moving-image design
• Build a custom trained ChatGPT assistant for concept ideation, story writing and directing prompts
• Generate and direct imagery in Midjourney: setup, parameters, prompt techniques, /describe, Style Tuner, character and style consistency
• Combine images experimentally in Artbreeder (Splicer, Mixer) to expand design thinking and produce fashion imagery
• Produce film concept art in Stable Diffusion: parameters, prompting, image-to-image, inpainting, upscale and ControlNet
• Plan a title sequence: ideation, theme/motif/symbol setup, lighting, point of view and sequence rhythm
• Retouch AI frames in Photoshop (including Generative Fill) and raise resolution while keeping detail
• Balance control and exploration so AI output serves your directing intent, not the other way round

✅ Requirements

• Skills: Basic design sensibility; no prior AI-tool experience needed (course level: basic to intermediate, theory taught before tools)
• Tools: ChatGPT, Midjourney, Artbreeder and Stable Diffusion accounts/installs, plus Photoshop for retouching (latest versions; programs not included with the class)
• Hardware: PC capable of running Stable Diffusion locally or via cloud, minimum 16GB RAM, dedicated GPU recommended

📝 Description

The class opens with what most AI courses skip: design theory. Through case studies of the “No Time to Die” and OFFF “Gemini” title sequences, Lee Heebok breaks down how abstract images gain depth and move people — fractal repetition, contrast, visual order and rhythm, the difference between visual and plot-based storytelling, and how theme, motif and symbol hold a sequence together at both macro and micro scale.

Then the tools. ChatGPT becomes a trained design assistant that absorbs your directing and produces concepts and stories. Midjourney covers the full span from installation and parameters to multi-prompts, image prompts, Style Tuner and consistency controls. Artbreeder trains combinational thinking through Splicer and Mixer experiments. Stable Diffusion goes deepest: parameters, prompting technique, image-to-image, inpainting, upscale and two full ControlNet chapters, closed with Photoshop retouching and Generative Fill.

Every block ends in a real exercise with 30+ visual artworks: a “Fusion of Dreams” music-video styleframe set, experimental fashion images, character and scene concept art for the fictional scenario “ECHO: Shadows of Perception”, and finally a complete film title-sequence styleframe package — ideation, production and revision. Class materials include the lecture PPT and a custom GPT, and lessons are updated as the fast-moving AI tools change.

Total: 41 lessons, 19 hours 50 minutes. Level: basic to intermediate. Card built from the manufacturer page via the Coloso dump pipeline.

🧑‍🎓 Who this course is for

• Aspiring directors who want solid design principles and deeper storytelling from an industry top-tier veteran
• Motion graphic designers who want fast AI visualization so they can spend time on final output instead of test drafts
• Job seekers who want standout ability, from solid design theory to the latest AI techniques

🧑‍🏫 About the Author

Lee Heebok (이희복) is a creative director with 23 years in Hollywood, currently Creative Director at the LA branch of Giantstep, formerly at Method Studios, Prologue Films and Ivanist. His title-sequence credits include Speed Racer, Spider-Man 3, Superman Returns, Unknown and Ninja Assassin, plus the OFFF “Gemini” titles and an homage to the “Skyfall” sequence. His teaching motto — “quantity leads to quality” — frames AI as a way for designers to generate, test and refine more ideas faster, and this Coloso-exclusive class is the first to publish his full title-sequence process.

🏁 Final Result

• A complete fictional film title-sequence styleframe set — from ChatGPT ideation to Midjourney, Artbreeder and Stable Diffusion frames with Photoshop revision
• A reusable directing system: design principles, a custom GPT assistant, prompt technique and a control-vs-exploration balance for any AI design task

📢 Guys, this is a Group Buy campaign, check the GBC forum thread to get access

You can find the course preview in the download section below.