AI has already created tremendous opportunities for designers working on traditional 2D applications and services. But what has excited me even more is how AI is fundamentally changing the way we design spatial computing experiences. The reason is simple: Spatial computing has always had a much higher barrier to entry. Unlike traditional 2D interaction design, even simple interactions require mathematics, physics, and engineering.
I still remember the early HoloLens days when getting a simple cube to respond to gaze and air tap could easily take days. Behind that seemingly simple interaction were raycasts, hit testing, cursor projection, gesture recognition, and visual feedback, all requiring significant engineering effort.
Frameworks such as MRTK and Meta XR Interaction SDK dramatically lowered that barrier by providing reusable interaction building blocks. But refining an interaction until it actually felt natural still required deep technical knowledge.
This is where AI has become a game changer. Over the past year, I’ve been experiencing this shift firsthand, both professionally and through my personal project Cosmic XR (https://lnkd.in/g-QYD7rw). While building Cosmic XR, AI helped me achieve complex interactions such as natural object manipulation with realistic inertia while exposing the right tuning parameters. That allowed me to focus on refining the experience instead of building the underlying plumbing.
AI doesn’t replace mature SDKs. It amplifies what designers can accomplish on top of them. Instead of spending hours figuring out implementation details, we can prototype ideas, generate interaction logic, and iterate on what ultimately matters: how the interaction feels.
Just as AI transformed 2D design workflows, it is now reshaping how we prototype spatial experiences. The challenge is no longer simply knowing how to implement something. It’s defining the right problem and communicating the desired experience spatially. Multimodal AI makes this even more powerful. A simple sketch often communicates spatial relationships and interaction flow better than several paragraphs of text prompts.
As AI continues to evolve, I believe one of the most valuable skills for XR designers will be spatial thinking: understanding human behavior in three-dimensional space and communicating those ideas clearly enough for both humans and AI. I explore many of these ideas in my new book, Spatial Computing: XR Interaction Design. (https://lnkd.in/g7_dHAPv)
Design taste and judgment become even more valuable, especially in spatial computing. Making an interaction feel right isn’t about solving a single technical problem. It’s the result of getting countless subtle details right, including spatial relationships, timing, animation, visual and audio feedback, affordances, interaction thresholds, hysteresis, and false positive prevention. Most of these are things users rarely notice consciously, but immediately feel when something is off.
English 
Korean