Meta Connect 2025: Ray-Ban display glasses at $799, failed live demo, and the road to AR dominance

Summary

{
  "long_summary": "Meta announced Ray-Ban display glasses at $799 and a neural wristband at its Connect conference, positioning itself ahead of Apple in the AR race. The company also demoed 3D Gaussian splatting on the Quest headset and showcased what it calls \"personal super intelligence\"—agent-driven commerce that will let users photograph products and order them directly through the glasses.\n\nThe headline moment was a live WhatsApp video call demo that failed on stage. The hosts, who had tested the glasses extensively in advance, attribute the failure to simultaneous livestreaming load rather than a product flaw. They note the demo worked reliably during private testing. The failure itself became a secondary narrative: several industry observers, including Ben Thompson and Dylan Field (Figma CEO), praised Meta for taking the live-demo risk, contrasting it favorably with Apple's reliance on prerecorded demos.\n\n**Hardware positioning and subsidy question**\n\nAt $799 for display glasses weighing just 69 grams, Meta is undercutting the original Google Glass price ($1,500) by half. The hosts debate whether Meta is subsidizing the hardware. Given that the Orion prototype reportedly cost $10,000 to produce, the $799 price seems remarkably low. But Meta's scale in hardware (Ray-Ban sales in the millions, Instagram partnerships driving billions in fashion advertising) and willingness to absorb losses on a limited initial run make subsidization plausible. Meta doesn't need to profit on hardware; owning the next computing platform is the strategic bet.\n\n**Apple's timeline and competitive dynamics**\n\nMark Gurman reports Apple's first non-display glasses are expected late 2026 to early 2027—meaning Meta has two to three years to iterate before facing a direct competitor. Apple's eventual display glasses won't arrive for \"a few more years\" after that. The hosts note that Apple's glasses will offer tighter Bluetooth integration and superior cameras, but lack Meta's platform advantages: Instagram, WhatsApp, and a consumer base already trained to share on Meta's networks. Meta AI is shipping sooner than Apple Intelligence or any AI assistant reliably accessible through voice on Apple hardware. Siri integration with ChatGPT remains clunky enough that users still open the ChatGPT app manually.\n\n**Product insights from hands-on testing**\n\nThe hosts tested the neural wristband extensively and found it intuitive—similar to iPhone gestures but adapted for wrist-based hand tracking. They see the real value in moving compute and battery off the face to the wrist, reducing strain on the bridge of the nose. Boz suggested the neural band could eventually become a platform input device for third-party developers. Speech-to-text is likely the dominant input method, with handwriting recognition filling a smaller niche for short responses.\n\nOn 3D Gaussian splatting, the use cases remain novelty-driven. The hosts see potential in real estate (scanning properties for remote touring) and as a pipeline to game development, but acknowledge the feature lacks immediate killer apps.\n\n**\"Personal super intelligence\" and commerce**\n\nMeta didn't detail this extensively, but the hosts infer it mirrors what Sam Altman outlined for ChatGPT: a shopping agent that lets users photograph or screenshot products and order them. Meta is unlikely to abandon its advertising relationships with brands; instead, it's building an agent-to-commerce layer that monetizes discovery. Integration with Shopify or direct brand APIs is probable. This is an extremely valuable revenue lever—essentially taking a cut of e-commerce transactions initiated through Meta platforms.\n\n**Organizational and thematic notes**\n\nThe Meta executive team is notably young (most in their 40s) with 15-year average tenure. That stability and long-term alignment shapes the confidence to take live-demo risks. The hosts also surface Meta's pattern of mythologically coded project names—Orion (the hunter struck down by hubris), Prometheus (stealing fire from gods), Behemoth (the untamable beast). Whether intentional or not, these names carry warnings alongside ambition. The discussion of Ava Chen, a 15-year Instagram veteran whose work on fashion partnerships is central to the platform's influence, underscores that executing on wearables requires taste and distribution, not just engineering.\n\nOverall assessment: Meta is swimming in an underhyped lane. While AI dominates attention, wearables—particularly low-cost, always-on, camera-and-AI-enabled glasses—are advancing faster than most observers expect. Apple's two-year gap before launching non-display glasses creates a window for Meta to build network effects and developer momentum."
}