Meta's Muse AI agent to be integrated into smart glasses
Meta's AI agent, Muse, will soon be available on smart glasses, allowing users to complete tasks via voice commands without needing their phones.
Meta announced that its AI agent, Muse, will be launched on smart glasses platforms in the near future. This integration will enable users to issue voice commands, and Muse will perform various tasks in the background without requiring the user to access their phone.
Muse operates on the Muse Spark AI model and runs within a private, persistent Linux virtual machine in the cloud. This environment includes its own web browser, file system, and terminal, which continue to operate even when applications are closed. The AI can write and execute code, delegate tasks to sub-agents, and schedule future actions. It is also capable of generating documents, PDFs, websites, and interactive tools like expense trackers.
Unlike AI agents primarily aimed at developers, Muse is designed for general consumers. It features a simplified, instant messaging-like interface with an avatar and highlights practical everyday use cases. Meta states Muse can assist with tasks such as booking travel, sending emails, filling out forms, and online purchases. It can also handle more complex requests like negotiating bills or facilitating car sales.
For smart glasses, potential applications include booking services, checking flight prices, finding replacement parts, logging meals, offering fitness advice, and assisting with purchases of visible items. Users will activate Muse with a voice command like "Hey Muse." Meta also addressed privacy concerns, stating that Muse data will not be used for advertising unless users opt-out, though it will be used for training future models. Meta plans to introduce a Confidential VM option later this year for enhanced privacy.