A person in a white VR headset reaching a motion controller toward the camera, lit in purple and pink light
A person in a white VR headset reaching a motion controller toward the camera, lit in purple and pink light

•

8 mins read time

Spatial UI Design for AI Glasses and Headsets: The Glance, Talk, Recede Test

Spatial UI design is less about 3D worlds and more about glances, voice and assistants. Here is what carries over from mobile, what breaks, and what to change in the next 12 months.

Ayoub Kada

•

8 mins read time

Spatial UI Design for AI Glasses and Headsets: The Glance, Talk, Recede Test

Spatial UI design is less about 3D worlds and more about glances, voice and assistants. Here is what carries over from mobile, what breaks, and what to change in the next 12 months.

Ayoub Kada

The headset demo is not the shift. The shift is your product being reached by a sentence, answered in a glance, and gone a second later.

Spatial UI design: what changes when the screen is the room

Spatial UI design is the work of placing interface in the world around a person instead of inside a rectangle they hold. For most product teams the honest answer to "should we prepare?" is yes, but not by building a headset app. Prepare by making your product work in three new conditions: a glance on a pair of glasses, a sentence spoken to an assistant, and a window floating in a room.

The platforms already say this plainly. Google's design principles for Android XR glasses tell teams to design for hands-free input such as voice or gesture and to make UI "appear only when needed and recede when the task is complete" (Android XR glasses design principles). Apple's guidance for Vision Pro asks designers to find "the minimum level of immersion" each moment needs (Designing for visionOS). Neither vendor is asking for spectacle. Both are asking for restraint.

The hype vs the reality

The hype says every app becomes a 3D world, screens disappear, and designers who cannot model in three dimensions are finished. That story has been told several times before, and it keeps colliding with the same facts: people wear headsets for focused sessions, not all day, and glasses with a small display or no display at all are a very different medium from a headset.

What holds up is quieter and more useful. AI assistants are becoming the main way people reach features on these devices, because typing on a face is not an option. That means the spatial shift is mostly an AI shift. Your product is reached through a spoken request, answered with a short glance or a sentence of audio, and only sometimes opened as a full window. If you have read the DIGCY piece on apps without screens, this is the same direction applied to hardware you wear.

Verdict: the 3D metaverse version is overhyped. The glanceable, voice-first, assistant-mediated version is underhyped, and it is the one that will reach most of your users first.

Mobile patterns vs spatial patterns

Some of what you know carries over. A lot does not. This table is the quickest way to audit a product.

Design decision

Mobile today

Display glasses

Headset (windowed)

Primary input

Touch

Voice, simple gestures, assistant

Eyes plus a pinch, voice

Session length

Minutes

Seconds

Focused sessions

Information density

Medium to high

Very low, one idea at a time

Medium, like a desktop window

Notifications

Banners and badges

Rare, never blocking the view

Contained to the window

Background

Your own pixels

The real world, always moving

Passthrough or an environment

Navigation

Tabs and stacks

Almost none, task in and out

Windows the user places

Brand expression

Visual identity

Voice, tone, timing, sound

Visual identity plus space and audio

Failure mode

Cluttered screen

Distraction while walking

Discomfort, neck strain, motion

Read the glasses column twice. It is where most mobile habits break. Persistent tab bars, dense cards, carousels and anything that asks for a long read while moving all fail there. Google's guidance specifically warns against expecting people to read large amounts of text while walking and against sudden notifications that obstruct the view.

The Glance, Talk, Recede test

A simple framework for any feature you want to make spatial-ready. Run each core task through three questions.

Glance. Can the essential result be understood in one look, roughly a single line or a single number? If the answer needs a paragraph, it belongs on the phone or in a window, and the glasses version should hand off rather than cram.

Talk. Can the task be started by a natural sentence to an assistant, without the user knowing your navigation? Your product's actions need clean names, clear parameters and confirmable outcomes. It is the same work that makes a product usable by AI agents, which is why multimodal UX and agent readiness keep showing up in the same planning meetings.

Recede. Does the interface leave once the job is done? On a phone, lingering UI is a nuisance. On glasses, it is a safety problem. Design the exit as deliberately as the entrance.

A feature that passes all three is ready for glasses and assistants. A feature that passes only Glance and Recede may still be a good headset window. A feature that passes none should stay on the phone, and that is a fine answer.

What product teams should do in the next 12 months

You do not need a spatial roadmap. You need a few changes that pay off on every surface, including the phone you already ship.

  1. Name your actions. Write down the ten things users most often come to do, phrased as sentences a person would say. Make sure each one can be completed or confirmed without the full app. This is groundwork for assistants, agents and glasses at once.

  2. Design a one-line state for every core flow. Order status, next meeting, balance, step two of five. If you cannot express it in a line, glasses will expose that gap.

  3. Treat sound and timing as brand. On displayless glasses your brand is a voice, a tone and a sense of when to stay quiet. Motion teams already think in timing; extend that thinking to audio cues. Good motion design and AI video work already starts with that kind of timing decision.

  4. Build adaptive layouts, not fixed screens. Spatial windows can be resized and moved; glasses show a sliver. Components that adapt by density, as discussed in adaptive UI, travel better than pixel-perfect mockups.

  5. Write the privacy story into the UI. Cameras and microphones on faces raise the stakes. Google's guidance asks apps to be transparent about what data they collect, why, and how it is stored. Show it at the moment of capture, not in a settings page.

Where this does not apply: if your product is a dense professional tool used at a desk for hours, glasses will not change much for you soon. A headset window might, but treat it as an experiment, not a migration.

Where design goes over the next one to three years

Expect the job to split less by device and more by moment. Someone will own the glance, the spoken request and the handoff between surfaces. That person is still a product designer, but their deliverables look different: intent maps, state lines, audio and motion specs, and rules for when the AI speaks versus when it shows.

The rectangle does not die. Phones and laptops stay where people do deep work. What changes is that the rectangle stops being the only place your product lives, and the AI assistant becomes the front door to many of its features.

The fear: is my skill set obsolete?

Most designers asking this are not worried about 3D modeling. They are worried that a decade of screen craft will not count. Here is the plain answer: the hardest parts of spatial UI design are hierarchy, restraint, legibility, timing and trust. Those are screen skills, applied under tighter limits. The designers who struggle will be the ones who equate design with filling a canvas.

The skills worth building now are voice and conversation design, motion and audio timing, writing tight microcopy, and working with engineers on structured actions that assistants can call. None of these require a headset on your desk. If you want a broader take on the replacement worry, read will AI replace UX designers.

FAQ

What is spatial UI design?

Spatial UI design is the practice of designing interfaces that exist in a person's physical surroundings, on headsets, glasses or ambient devices, instead of on a flat screen they hold. It covers placement in space, input by eyes, hands and voice, glanceable information and comfort.

Do product teams need to build a headset app now?

Usually not. Most teams get more value from making core tasks work through voice, assistants and one-line states. Those changes help on glasses, headsets and phones at the same time. Build a headset app only when a task truly benefits from space or immersion.

How is designing for AI glasses different from designing for a phone?

Glasses favour seconds, not minutes. Input is mostly voice or simple gestures, information must fit a glance, the background is the moving real world, and UI should disappear when the task ends. Google's Android XR guidance also covers glasses with no display, where audio carries the whole experience.

Will spatial computing replace UI designers?

No. It changes what they deliver. Hierarchy, legibility, restraint and trust matter more on a small transparent display than on a phone. Designers who add conversation design, motion and audio timing to their screen craft will be in a stronger position, not a weaker one.

Is spatial UI overhyped?

The immersive 3D-everything version is overhyped for most products. The practical version, short glanceable answers and assistant-driven actions, is real and arriving through AI glasses and system assistants. Plan for that version first.

Getting ready without betting the roadmap

The cheapest way to prepare for spatial interfaces is to make your current product clearer: named actions, short states, adaptive components and a privacy story people can see. If you want help turning that into a product plan, a UI/UX agency in Casablanca like DIGCY works with teams globally on AI product development and product design services for exactly this kind of multi-surface work. DIGCY is a Dribbble Selected Agency, and its ZeroDrift work received an Awwwards Honorable Mention. Start with the Glance, Talk, Recede test on your top five tasks. The results will tell you more than any headset demo.

Let’s keep in touch.

Discover more about high-performance web design. Follow us on Twitter and Instagram.

Building an AI product?

By submitting, you agree to our Terms and Privacy Policy.

Let's Talk

Tell us about your AI product. After you send this, you can pick a 20-minute slot on our calendar. Prefer email? hello@digcy.com We reply within one business day.

Building an AI product?

By submitting, you agree to our Terms and Privacy Policy.

Let's Talk

Tell us about your AI product. After you send this, you can pick a 20-minute slot on our calendar. Prefer email? hello@digcy.com We reply within one business day.

Building an AI product?

By submitting, you agree to our Terms and Privacy Policy.

Let's Talk

Tell us about your AI product. After you send this, you can pick a 20-minute slot on our calendar. Prefer email? hello@digcy.com We reply within one business day.