OpenAI's upcoming hardware family, including a smart speaker and glasses, will intentionally have no screens. This is a deliberate strategic choice to move beyond the screen-centric ecosystem dominated by Apple and Google. It represents a bet on a future where AI interaction is primarily ambient, powered by voice and computer vision rather than touchscreens.

A single, context-aware AI assistant with access to various APIs will replace dozens of specialized apps for tasks like fitness tracking, to-do lists, or flight check-ins. Users will interact conversationally with their assistant, rendering most single-purpose apps redundant.

Context-aware personal agents will subsume the functions of many standalone apps, such as fitness or calorie trackers. An agent that already knows a user's location, schedule, and goals can perform these tasks more seamlessly, reducing many current apps to mere APIs for the agent to consume.

The ultimate winner in the AI race may not be the most advanced model, but the most seamless, low-friction user interface. Since most queries are simple, the battle is shifting to hardware that is 'closest to the person's face,' like glasses or ambient devices, where distribution is king.

The ultimate vision for AI wearables isn't just a voice-activated search engine, but a contextually-aware assistant. It will proactively offer help based on your situation—like suggesting what to do next after a conversation ends—without explicit commands, becoming a true partner rather than just a tool.

Demis Hassabis suggests that previous attempts at smart glasses like Google Glass were too early because they lacked a compelling use case. He believes a hands-free, always-on AI assistant like Project Astra provides the 'killer app' that will finally make smart glasses a mainstream consumer device.

Adding existing health sensors like heart rate monitors to new devices like smart glasses offers diminishing returns. The real innovation and value proposition for new wearables lies in developing new interaction paradigms, particularly advanced, low-latency audio interfaces for seamless communication in any environment.

Demis Hassabis claims previous smart glasses failed not just due to clunky hardware but because they lacked a compelling use case. He argues that a powerful, seamless AI assistant, integrated into daily life, is the "killer app" that will finally drive adoption for this form factor.

The compelling feature for future AI wearables is persistent audio recording and synthesis. The ability to listen to your day and automatically generate tasks and summaries is a "holy crap moment" that will make today's notification-centric smartwatches seem primitive by comparison.