Meta’s annual show is set to move the focus from headsets to wearables, smart glasses and on-device AI will be its stars. Look for a push toward display-sporting eyewear, an ultimate neural wristband control solution and a wider strategy around Horizon OS to entice developers — and sooner or later, third party hardware – in.
Ray-Ban smart glasses go from camera-first to AI-first
Meta will need to be careful not to screw up too much when refreshing its Ray-Ban lineup. The existing models already offer a 12MP camera, hands-free capture and multimodal Meta AI for tasks such as live translation and object recognition. The next version is likely to address longer battery life, improved video quality and better audio pickup in noisy city streets — key drawbacks for everyday use.
A long-sought change would be to shoot video in a native landscape mode instead of going vertical by default, aligning the glasses with creators who post across platforms. Industry buzz also hints at two separate versions, one made for wearing as outdoor sunglasses and the other intended as optical lenses, a maneuver that would expand appeal to all-day wear. The existing models which are based on Qualcomm’s AR-class silicon; it would also help to unleash faster on-glass AI prompts and lower-latency voice interactions.
Hypernova: display-enabled glasses, come through
The headline grabber will likely be a new version of its Meta smart glasses with built-in waveguide display—widely assumed to come out under the Hypernova program, though any retail branding is yet to be confirmed. Imagine glanceable, heads-up overlays for navigation, messaging triage and contextual prompts that surface without a phone needing to be raised. There have been a number of early-industry prototypes from various players that suggest the MicroLED powered waveguides can work, but field of view and brightness remain difficult trade-offs – Meta’s implementation will be watched to see how it stacks up.
If Meta holds to the formula it’s been iterating on, you can expect a mostly practical ride — some speedy turn-by-turn arrows, notes left at places and AI that smartly highlights what’s relevant right now.
Privacy indicators — visible LEDs for camera or capture — are a virtual certainty, based on regulatory and public expectations that have been set by previous wearables from companies such as Snap and Google.
A neural wristband that makes the small gestures useful
Anticipate the first glimpse of a Meta wristband controller, which has been boiling since its CTRL-Labs acquisition. The as yet rumor- and codenamed-controlled device reads small electrical signals of intent from something worn on the wrist, such as a twist, flick or even scroll — minus any large, overt gesture. In theory, if the reality headset holds up in shipping hardware the sub-100ms latency lab-tested electromyography inputs — those could translate into “text input, selection and media control to feel completely like second nature” on the go, Reality Labs has suggested.
We’ve seen glimpses of what’s possible with third-party EMG accessories like Wearable Devices’ Mudra and research out of university labs. The key here is tight integration: A neural band coupled with Meta’s glasses could finally provide wearable computing with a control scheme that is silent, precise, and socially acceptable.
Horizon OS and an inviting SDK
The biggest news might have been software, not hardware. And Meta is expected to open up an SDK for smart glasses apps, expanding its ambitions for the Horizon OS beyond VR applications and into “heads-up” use cases. That means actual APIs for notifications, navigation, fitness, translation and enterprise tools — not just to mention monetization and distribution paths that don’t depend exclusively on experimental betas.
Meta has said that it plans to have Horizon OS partners, and has already named ASUS’s Republic of Gamers and Lenovo for future devices. A public glasses SDK would provide developers a target and to foster design for micro-interactions; 2–3 second glances, voice and EMG input, and context-aware prompts. Llama 3–series models can be used to drive the same feature directly on-device (with cloud assist for heavier lifting).
VR takes a break as partners step in
This time, don’t look for a new Quest.
Instead look for third-party hardware momentum around Horizon OS. The chatter continues about an ASUS ROG headset — described as a performance-first device with eye and face tracking, plus high-resolution displays — set to demonstrate how the Horizon OS can reach beyond gaming into creator workflows. Another probable pillar is Lenovo itself, a sturdy name in productivity-first mixed reality.
This pivot is in line with market dynamics. I’ve always been a static VR evangelist, in private if not in print (who even knows how it looks behind those closed windows?). Analysts at the industry research firms IDC and CCS Insight have both observed that although VR is just as likely to spike according to major product releases, everyday wearables feed steadier engagement. A larger OS with many OEMs spreads risk and speeds up experimentation.
Horizon Worlds and Meta AI become more generative
On the platform side, expect generative creation in Horizon Worlds — text-to-scene creating, physics-aware props and character agents with adjustable personalities. Filled with too much gravity, the idea of an Llama-styled take on building out user-generated worlds has a simple barrier to entry and fast iteration loop that gives creators far more tools to work with than we’ve seen from Meta since slipping into its Ray-Ban-clad video and photo clairvoyance.
Meta also seems poised to push character-driven assistants into new languages and regions, a crucial development for extending their global reach. The question will be how to balance utility and safety — guardrails, cues of transparency and on-device processing options will all be scrutinized by regulators and users alike.
Why this Connect matters
If rumors pan out then Meta will usher in the next era of consumer AR — from camera glasses to glanceable displays, voice-only control to EMG, and closed features to a real developer platform. That combination is what transformed smartphones from gadgets into ecosystems. That exact playbook — enacted with privacy, battery efficiency and developer ROI in mind — could come to define the next decade of spatial computing.