Wouldn't it be cool? - An AI Generated Climate Change Series

Inspiration

One of the first things I did when I started playing around with AI was probing the thousands of pages of the IPCC reports. These are some of the densest but most critical publications of our time and I found so many wonderful gems buried in the data. These reports contain real, tangible solutions and pathways to a future not yet realized, but they're written in negotiated phrases and obscure jargon that few people will ever read. I wanted to find those hidden gems and bring those possibilities to life.

What it does

"Wouldn't it be cool?" discovers optimistic, real-world climate solutions people are creating to adapt to our changing climate. The series makes complex climate science accessible through compelling visual storytelling powered by generative AI. Episode 1 explores why optimism about climate is difficult, explains the concept of "Net Zero" using an engaging bathtub analogy, and begins imagining the world we want to build in the face of climate change. The series focuses on solutions that don't just reduce carbon in our daily lives, but also make the future better than it is today.

How we built it

The research phase used NotebookLM to query IPCC reports and climate data, with Claude and Gemini helping to refine language and phrasing. I wrote scripts in a three-column format (script, visual description, and storyboards generated in Firefly boards.) and visuals side-by-side) For visual assets, I generated key start and end frames using Freepik's unlimited image generation (primarily with Seedream 4k and Google's Nano Banana). I used ComfyUI workflows for inpainting with Qwen Image edit and Photoshop's generative fill for outpainting. I then upscaled everything to 4K using Magnifik.

After recording the main voiceover, I created an animatic rough cut with 4K images and audio elements generated through Suno (music), the Adobe Audio Library, and ElevenLabs (SFX and additional voices). This animatic helped establish timing for video generation.

For video generation I used Google Veo 3.1 (via Freepik) for lip-sync dialogue and Kling 2.1, Seedance or Wan 2.2 for other shots depending on motion needs. After swapping clips into the animatic for final timing, I upscaled the video to 4K using Topaz Video. Motion graphics were created in After Effects, with final assembly, color grading, and audio mixing completed in Premiere Pro using conventional techniques.

Challenges we ran into

Control and consistency are still regular challenges when working with generative AI. In the end, I decided to embrace some of the idiosyncrasies of AI as part of the transparent process of using these tools to help educate and inspire. Rather than fighting for pixel-perfect consistency, I leaned into the experimental nature of the medium and treated this series as both a climate story and an open exploration of how we tell stories with emerging technology.

Accomplishments that we're proud of

• Creating a climate documentary that focuses on optimism rather than doom while maintaining scientific accuracy.
• Successfully mining thousands of pages of IPCC reports to surface compelling, real-world solutions that typically remain buried in dense jargon.
• Demonstrating that complex topics like climate science can be conveyed with generative AI in a way that is genuinely accessible and engaging for broad audiences.

What we learned

I've been producing and editing for more than two decades, and the process of developing this project had me reimagine my entire creative workflow, from pre-visualization and storyboarding to the final edit. Instead of being constrained by budget limitations that would have meant simple 2D animations, generative AI opened up far more compelling visual options. I discovered I could keep image generation tools open while writing, instantly testing visual ideas and styles that could later become references. The performances from the video generations were often surprising in delightful ways and I would rewrite the narration to accomodate those moments. In the end I was finding ways to creatively iterate in ways that were never possible before.

What's next for Wouldn't it be cool? - An AI Generated Climate Change Series

Future episodes will tackle the complex challenges facing energy production, agriculture, construction, transportation, and more. We'll continue to focus on solutions that don't just reduce the carbon in our daily lives, but also make the future better than it is today. Each episode will explore different sectors while continuing to experiment with generative AI as a storytelling tool, pushing the boundaries of how we can make dense scientific reports and complex climate solutions accessible, engaging, and ultimately inspiring for audiences worldwide.

Built With

• adobe
• claude
• comfyui
• elevenlabs
• freepik
• gemini
• kling
• magnifik
• nanobanana
• notebooklm
• qwenimageedit
• seedream
• suno
• topaz
• veo3.1
• wan2.2