New Categories, Personalized AI, and Earbud Cameras
The Qualcomm Analyst Relations team recently organized a chat for me with Dino Bekis, VP and GM of Wearables at Qualcomm. Having met Dino several times, I have always found our conversations insightful. My goal for this session was to understand the current roadmap and where the industry is heading.
Two major themes stood out. While the future certainly includes better smartwatch chips, enhanced compute power, and improved wireless audio, the truly exciting frontier involves wearable AI that knows you and audio devices equipped with low-power, low-resolution, AI-capable cameras.
Biometrics and Fingerprint Technology
Dino’s group also oversees the ultrasonic fingerprint business. I was able to confirm the significant performance leap Google achieved moving from the Pixel 8 Pro to the Pixel 9 Pro. The detection is noticeably more accurate and smoother, and the Pixel 10 Pro XL has further improved ease of use. While screen protectors remain a challenge for ultrasonic tech, it is still a significant advantage over optical sensors, which are affected far more severely by third-party glass.
The Smartwatch Landscape
Qualcomm is performing exceptionally well in the smartwatch space. Outside of Samsung, nearly every major player in the Wear OS ecosystem utilizes the Snapdragon. The latest iteration, the Snapdragon W5 Gen 2, powers the Pixel Watch 4 (both 41mm and 45mm versions, including Wi-Fi and LTE models). Partners like OnePlus, Xiaomi, Oppo, and Vivo remain strong, bringing this technology to market at more accessible price points.
The Gen 2 platform now supports satellite connectivity, potentially offering life-saving features for outdoor enthusiasts. This new chipset, which has shrunk from a 12 nm process with Wear 4100 to 4nm with Wear 5100 Gen 2, provides a significant boost in performance while helping Google and others increase battery life. Moving forward, Snapdragon Wearable platforms are destined to become even more power-efficient, with Qualcomm focusing on “Distributed AI”—running workloads on the most efficient hardware available.
Miniaturization and New Form Factors
Dino expressed satisfaction with the progress made in wearables over the last decade. While I argued that XR glasses haven’t reached the mainstream as quickly as expected, Dino highlighted the ongoing challenges of form factor, power consumption, and fashion. Meta’s partnership with Luxottica (Ray-Ban) proved that consumers prioritize fashionable aesthetics over pure technology specs.
They are managing customers developing audio-only glasses that feature connectivity and AI/machine learning capabilities. I can easily imagine some of these eventually incorporating cameras. Dino also mentioned the Meta wristband as a joint platform: the wearables team manages the wristband hardware, while the XR team manages the glasses.
The Rise of “AI That Knows You”
Customers are currently experimenting with the definition of what a “wearable” actually is. Whether it’s a device in a pocket, a necklace, or a pin embedded in clothing, Dino anticipates an explosion of new categories.
There is increased interest from non-traditional players—specifically cloud providers and model developers—who want to create AI appliances. The primary goal is to integrate these capabilities into a personalized experience. Imagine an AI that truly knows you by utilizing a mix of sensor data and location—all kept private—to provide context-aware, agent-based interactions.
Currently, if you ask ChatGPT a generic question, you get a broad answer. If the model knows who is asking and has the context to build a better prompt, the result is significantly better. This solves a major pain point: the need to constantly refine prompts or repeat details to get a useful response.
AI in the Right Place
A key Qualcomm philosophy is “AI in the right spot.” Your XR glasses or watch might send a request to the phone in your pocket, or even a nearby PC, to process a complex task before returning the result to the wearable. This seamless handoff between devices is something the company has been perfecting for years.
The Future: Cameras in Your Ears?
The detail that most caught my imagination was the prospect of earbuds equipped with low-power, low-resolution cameras.
While currently in the research phase, one can imagine a camera capable of running small AI models for object and context detection. This could enable real-time translation or navigation assistance. This concept is a major selling point for XR glasses like Meta Ray-Bans, but not everyone wants to wear glasses.
Consider the Total Addressable Market (TAM) for contact lenses—a multi-billion dollar industry catering to people who prefer not to wear frames. This group represents a massive demographic of early adopters who might prefer AI-powered earbuds to learn about their environment without needing to put on a pair of glasses.
