Source: Unsplash
Two smart glasses product launches. One week apart. One priced at $299 and one at $2,195. Both are betting their companies on getting wearable AI right. Both represent completely opposite UX philosophies. And only one of them understands what actually makes a wearable work.
Today, Meta officially launched its own-branded AI smart glasses starting at $299, the first time the company has put its own name on a glasses product rather than partnering with Ray-Ban or Oakley. Last week, Snap launched its long-awaited Specs AR glasses at $2,195 with full augmented reality overlays, a 51-degree field of view, and dual Snapdragon processors. The tech press is covering this as a competition. I want to look at it differently. This is a story about two competing UX philosophies for what wearable AI should actually feel like, and one of these philosophies has already killed a product before.
"Now, in the AI world, the things these glasses can do to help you out without constantly distracting you — that capability is much higher."
— Sergey Brin, Google Co-Founder, Google I/O, May 2025
The Numbers Say Meta Has Already Figured Something Out
Before we get into design philosophy, let's look at what the market is actually telling us. Meta's Ray-Ban smart glasses sold over 7 million units in 2025, tripling from roughly 2 million combined in 2023 and 2024. EssilorLuxottica, Meta's manufacturing partner, was already planning to scale production to 10 million units by end of 2026, and reports now suggest that target is being pushed to 20 to 30 million.
Meta's Ray-Ban glasses don't do true AR. There's no digital overlay on your view. They add audio, an AI assistant, and a camera. That's it. And yet they're the most successful smart glasses in history. Meanwhile, every product that promised true AR has either folded, pivoted to enterprise, or is just getting started at a price point that limits it to a niche.
That gap should be telling product teams something important. The technology people are actually adopting isn't the most impressive technology. It's the most invisible technology.
What Google Glass Actually Got Wrong
Sergey Brin stood up at Google I/O in 2025 and said something that the industry needed to hear: "I made a lot of mistakes with Google Glass." The product launched in 2013 at $1,500. It lasted barely a year in public before the "glasshole" backlash killed it. The commonly told story is that Glass was ahead of its time. I disagree with that framing. Glass was wrong in its time, in ways that were entirely predictable if you approached it as a UX problem instead of an engineering achievement.
The failures weren't primarily technical. The camera light that blinked when recording made everyone around the wearer uncomfortable. The tiny prism display demanded active attention, pulling your eyes away from the conversation you were having. The interface required deliberate interaction: head tilts, voice commands, taps on the side. Nothing about Glass felt passive or ambient. It was a device that constantly reminded you, and everyone around you, that it was there.
That is the opposite of what a wearable should do. A wearable interface that demands attention is a wearable interface that will be removed and left on the nightstand.
- The social UX problem: Wearables don't just have to work for the person wearing them. They have to be socially acceptable to everyone else in the room. Google Glass failed this test completely. The camera, the glowing display, the tilting interactions: all of it signaled "this person is recording you and doing something weird."
- The attention demand problem: Glass required you to shift focus from the physical world to the digital one. Every interaction broke your ambient awareness. A device you wear on your face should reduce cognitive load, not add to it.
- The always-on anxiety: Having a recording device on your face 24/7 made people around you uncomfortable. The UX of a wearable extends to the people who didn't choose to wear it.
- The 4-hour battery as a design constraint: Glass had a 4-hour battery. Snap Specs in 2026 also have a 4-hour battery. When your device dies mid-afternoon, you've designed a product that's only half a day useful. That's not a technical limitation: it's a UX constraint that demands a completely different interaction model.
- The "why am I wearing this" problem: Most Glass users couldn't answer this question confidently. The product had features, but it lacked a clear, daily-habit-level use case. Every successful wearable has one reason people put it on every morning. Glass had fifteen weak ones.
Snap's $2,195 Bet and the True AR Challenge
Snap's Specs are genuinely impressive on paper. 51-degree field of view. Full color digital overlay. Dual Qualcomm Snapdragon processors. Electrochromic lenses that shift from clear to sunglasses in 10 seconds. This is the first consumer AR product that gets close to what science fiction promised. And Snap's stock dropped nearly 10 percent on announcement day.
Here's what makes me cautious about Specs, even though I think the hardware is remarkable. True AR creates a harder UX problem than audio-only glasses, because now the interface literally competes with your vision. Every time you show someone a notification, a map overlay, or a Snap story floating in their field of view, you're asking them to split their attention between the physical world and the digital layer. That split-attention design is extraordinarily difficult to get right.
The companies that have done this well in adjacent fields, aviation HUDs, industrial AR, surgical guidance systems, all built their interfaces around one core principle: show less, not more. The instinct for new AR platforms is always to demonstrate capability by showing everything. The instinct that actually serves users is to show almost nothing, only when it genuinely helps, and then get out of the way immediately.
The Context Problem: When Should Glasses Show You Anything?
This is the design question nobody in the smart glasses industry is answering publicly, and it's the most important one. When should the glasses surface information? When shouldn't they? How does the system know the difference?
Think about notification design on a smartphone. We've spent fifteen years getting that wrong: app badges, banner notifications, full-screen takeovers, vibrations at 2am. We still haven't fully solved it. Now imagine that notification surface is directly in front of your eyes. There is no more intimate interface. There is no higher-stakes notification design problem in the history of computing.
The AR glasses that will actually succeed are not the ones with the most impressive displays. They're the ones that have thought through context-aware information delivery: when you're in a meeting, show nothing. When you're walking to an unfamiliar destination, show turn-by-turn navigation that disappears when you arrive. When you're speaking to someone in a language you don't know, show a subtle translation. Never all three at once. Never when the moment doesn't call for it.
This is not a hardware engineering problem. It's a product design problem. And it requires exactly the kind of thinking that most AR hardware companies don't have in their leadership teams right now.
I wrote about a related challenge on reloadux when talking about AI product readiness: the biggest barrier to AI adoption in enterprise isn't the model quality. It's whether the product team understands when not to use AI. The same principle applies here, at a much more intimate scale.
Meta's Design Philosophy: The Interface That Disappears
Meta's $299 glasses represent a deliberately constrained design choice. Audio only. Camera optional. No display. The AI assistant responds to your voice and speaks into your ears. To everyone around you, you're just wearing glasses. There's nothing to see, no camera light blinking, no glowing lens, no head tilts or finger taps.
What Meta figured out, and what the 7 million unit sales confirm, is that the best wearable AI is the one that feels like an upgrade to a thing you already wear. Not a gadget strapped to your head. An upgrade to your glasses. The interface disappears because the object it's attached to is already invisible in social contexts. We've been wearing glasses for 700 years. Nobody thinks twice about them.
The $299 Kylie Jenner collaboration edition is not a gimmick. It's a statement about what Meta thinks wins this market: fashion credibility and social normalcy before technology ambition. That might feel like a concession. I think it's the correct read of where mainstream users actually are.
What This Means for Designers Building for Spatial Computing
If you're a product designer and you're not already thinking about how your product might surface on AR glasses, you're behind. The platform shift is slower than the press cycle suggests, but it's real. Here's what I'd take from this week's news:
Design for ambient attention, not active attention. The dominant smartphone interaction model assumes a user who picked up their phone, opened an app, and chose to engage. AR glasses will have a dominant interaction model built around users who are in the middle of doing something else. Your interface needs to work in the periphery of their attention, not at the center of it.
The form factor is the UX. Snap's Specs will tell us in 12 months whether consumers are willing to wear something that looks slightly unusual in exchange for true AR capability. My bet is that the 4-hour battery and the $2,195 price are not the main barriers. The main barrier is whether people feel comfortable wearing them in a coffee shop, at a dinner, in a meeting. If the answer is no, the specs don't matter.
Think about social UX, not just user UX. When you design for a wearable device, your users are also the people standing next to your users. If the device makes bystanders uncomfortable, self-conscious, or suspicious, your user will stop wearing it regardless of how much they personally love the features. This is the design principle that killed Google Glass and it hasn't gone away.
There's a version of smart glasses that genuinely improves daily life: context-aware, socially invisible, useful in the exact right moments and absent in all others. The product that gets that interface right will be bigger than the smartphone. The question is whether anyone in the current crop of hardware companies has the design discipline to build it.
What do you think? Is true AR the future, or will audio-first glasses like Meta's dominate the next decade? I'm genuinely curious what people building products think about this. Drop your take in the comments.
Sources:
1. CNBC — Meta announces new smart glasses starting at $299, as Zuckerberg keeps pushing wearables (June 23, 2026) — https://www.cnbc.com/2026/06/23/meta-glasses-are-new-smart-glasses-starting-at-299.html
2. TechCrunch — Snap finally debuts its long-awaited AR glasses, Specs, and, oof, they aren't cheap (June 16, 2026) — https://techcrunch.com/2026/06/16/snap-finally-debuts-its-long-awaited-ar-glasses-specs-and-oof-they-arent-cheap/
3. Road to VR — Meta Sold Over 7 Million Smart Glasses Last Year, Effectively Tripling Sales in 2025 — https://roadtovr.com/meta-ray-ban-smart-glasses-sales-tripled-2025/
4. TechCrunch — Google's Sergey Brin: 'I made a lot of mistakes with Google Glass' (May 2025) — https://techcrunch.com/2025/05/20/googles-sergey-brin-i-made-a-lot-of-mistakes-with-google-glass/
5. UC Today — Meta Unveils $299 AI Smart Glasses to Bring Wearable AI to the Mass Market — https://www.uctoday.com/immersive-workplace-xr-tech/meta-ai-smart-glasses-299-launch-2026/