The world stays in view. The overlay must earn interruption.

Side project · Exploration · Open-eye AR · Spatial UX · 2026

Open

Full project

Active probe — what’s been worked through on interruption cost, spatial anchoring, and AI overlay trust. Sections update as the work develops.

TIERSide project · Exploration
DEVICEMentra Live
FOCUSOpen-eye AR · Spatial UX
STATUSEarly probe · 2026

Design probe for Mentra Live — open-eye glasses where the UI sits on top of the real world instead of replacing it. The work is interruption cost, spatial anchoring, and AI content the wearer can partially verify in scene.

Open-eye means attention is split by default. Any display element is an interruption — and the wearer can always look past whatever you put in front of them. That is safety and constraint. The design has to earn its place in the field of view and know when not to be there.

AI-generated content in a live scene behaves differently than on a screen. The user can verify some things directly (spatial, nearby) and none of others (data, history, context). Knowing which is which shapes what the model should offer and what it should withhold.

Outcomes.

Constraint

Attention is split by default — display must earn its place in the field of view.

Design problem

Trust when some facts are verifiable in-scene and others are model-supplied.

Still open

Spatial anchoring vs. distraction; gaze-and-voice interaction without a primary screen.

Paired probe

Even Realities G1 — glance-scale AI in a 20° window.

Essay: Glanceable trust