Skybox AI creates stunning 360° worlds—but sometimes it gives you cloud-mountains and camels the size of small buildings. This post looks at how learning designers can use these beautiful glitches with purpose, turning surreal scenes into moments of reflection, immersion, and meaningful decision-making.

Skybox is at its best when it's setting a mood, not building a blueprint. The lighting, landscapes, and overall vibe can feel breathtaking—right up until you notice the camel with four knees or a shop sign that says “SUBWAX.” But that’s part of the magic. It’s a tool for atmosphere, wonder, and exploration—not pixel-perfect realism.

In this post, I walk through my workflow using Skybox to upload and remix custom 360° scenes—then test how they behave once exported. I share what worked, what broke, and the small details (like navigation settings) that completely changed how the scenes felt. If you're experimenting with immersive tools or just curious about what's possible with generative 360° environments, this one's for you.

I’ve been experimenting with Skybox, a tool that generates 360° environments from text prompts, to see how it could support immersive, decision-based learning. It’s not perfect—but it opened up a lot of creative possibilities and helped me think differently about space, emotion, and learner choice.