At What Point Do We Stop Calling Smart Glasses Experimental?

Garry Williams, partner at Astral City, examines how wearables are moving interaction into our line of sight, and why brands need to rethink how they show up - 17/02/26

Meta and EssilorLuxottica have just announced that more than seven million Ray-Ban smart glasses were sold in 2025.

Seven million pairs.

At some point, we have to stop calling this “experimental.”

When hardware ships at that scale, behaviour shifts. Quietly at first. Then all at once.

The original iPhone sold roughly six million units in its first year, according to Apple’s reported sales figures at the time. It didn’t look like a revolution then either.

Now we’re seeing something similar take shape.

With Google, Samsung Electronics and Snap Inc. moving decisively into wearables, Meta & Luxottica scaling distribution, XREAL and RayNeo iterating on lightweight displays (along with others), and Apple redefining spatial computing with Apple Vision Pro, this stops being about devices.

It’s several technologies landing at once.

AI. Cameras. Lightweight optics. Spatial software.

And when that happens, the interface moves.

This Isn’t About Glasses

It’s about how we interact.

Desktop was fixed.

Mobile was portable.

Glasses are persistent.

Mobile required that you pull a screen from your pocket and commit attention.

Glasses remove that barrier.

You glance.

You speak.

You keep moving.

It feels simple. That’s the point

And when interaction gets easier, people use it more.

We aren’t inventing new habits. We’re relocating them.

Billions already talk to machines every day. Voice assistants are embedded in most smartphones. Smart speakers are ordinary household objects. Speaking to software no longer feels unusual.

Glasses simply move that intelligence from the hand into the field of view.

And when something sits that close to your senses, expectations change.

Audio First, But Not Audio Only

For now, most glasses are audio-led.

Microphones.

Open-ear speakers.

AI processing.

Minimal visual output.

But that limitation sharpens the design.

In an audio environment, brand expression becomes behavioural.

There’s no layout. No visual hierarchy. No screen doing the work for you.

Brand shows up through tone. Through timing. Through restraint.

Speaking only when it matters.

Remembering context without feeling invasive.

If the assistant talks at the wrong moment, it’s intrusion.

If it hesitates, it feels unreliable.

These devices sit close to your ears and eyes. Mistakes feel personal.

And that changes how you build these systems.

We’ve been testing this in live environments. When intelligence sits that close to someone’s senses, small mistakes feel big. Timing becomes design. Silence becomes design.

Micro-Moments Instead of Sessions

Mobile got us used to sitting inside apps.

Glasses compress interaction into moments.

You look at a product and ask about availability.

You glance at a departure board and ask about delays.

You explore a physical object and ask what you’re seeing.

Five seconds.

10 seconds.

Without breaking stride.

Natural language is the only control layer light enough to work inside those windows.

Typing doesn’t work mid-stride.

Menus don’t work mid-glance.

Conversation does.

Subtlety Is the Hard Part

Displays are coming. In some cases, they’re already here.

Small overlays.

Directional cues.

Contextual snippets.

Right now, simple.

As hardware improves, voice will combine with visual reinforcement. Information will anchor itself beside what you’re looking at.

But the challenge isn’t adding more.

It’s knowing when not to.

We’ve been building conversational systems that operate alongside physical spaces rather than inside screens. When AI shares the same physical context as the user, expectations rise fast. It has to understand what you’re looking at. It has to respond without interrupting flow. The right information, at the right moment. No more than that.

That’s not a technical problem. It’s a design one.

Response Is the Experience

Behind what feels like a simple exchange is a constant loop.

Listen.

Understand.

Respond.

In persistent interfaces, that loop never really switches off.

Latency becomes emotional.

Accuracy becomes trust.

When intelligence shares your visual context, users assume awareness. If it misreads that context, it doesn’t feel like a glitch. It feels inattentive.

Most conversational AI today is functional.

That’s not enough.

When AI lives in your field of view, response is the product.

Tone. Timing. Brevity. Context.

Those decisions define whether the experience feels helpful or intrusive.

Designing for What’s Coming

Seven million pairs isn’t saturation.

But it’s a signal. Enough to matter.

The interface is moving into our line of sight.

The people building for it now will decide what feels normal later.