The smart glasses pitch has been the same for a decade. This time it will be different. This time we have AI.
It is always this time.
The killer app is your lost remote
Billions of dollars in R&D, the full weight of Google's Android XR initiative, multimodal AI, always-on cameras, and the best answer the industry can offer is a pair of spectacles that function as a slightly creepy Find My network for your wallet.
The speakers were not dismissive. They genuinely argued that accumulating dozens of small conveniences like this might collectively justify wearing the devices all day.
That is not a product thesis. It is a prayer.
Hardware looking for software looking for a reason
Several Android XR devices are expected from eyewear brands including Warby Parker and Gentle Monster. Early versions will focus on audio. Display-equipped models remain further out.
The pattern is familiar. Ship the hardware with limited functionality. Promise the transformative features are coming. Hope the AI catches up before the consumer loses interest.
Google has shown experimental concepts where glasses use camera data and AI to recall where objects were placed. The key word is experimental. The technical hurdles between a concept demo and a product that works reliably in a kitchen full of clutter are vast.
The "sometimes device" problem
Consumers still treat smart glasses as occasional gadgets rather than something they wear all day. This is the fundamental obstacle and nobody has solved it.
For constant wear to make sense, the benefits have to outweigh the discomfort of walking around with cameras and sensors strapped to your face. That is a high bar in a world increasingly sensitive to surveillance and data collection.
The industry's answer is that AI will clear that bar. But when pressed on specifics, the examples are modest and the timelines are vague.
The honest assessment
Smart glasses will probably get good eventually. The convergence of lightweight hardware, multimodal AI and edge computing points in that direction.
Related reading
- Meta launches Muse Spark as Zuckerberg bets on proprietary AI
- Jury finds Meta and YouTube liable for engineering addiction in children
- Google bets on AI coaching and simpler hardware with Fitbit revival
But eventually is doing a lot of heavy lifting. Right now the category is where voice assistants were in 2015, full of potential and short on reasons to care.
The industry needs a breakthrough application. Finding your keys is not it.