Relaying Voice Commands Between AI Systems

The patent introduces a broadcasting mechanism as an intermediary that enables unknown users to be recognized by known users' devices. When a user speaks to a device they don't own, the device broadcasts the voice command and user identifying information to other devices in the network, which can then execute the command if they recognize the user. This intermediary broadcasting layer resolves the contradiction by allowing the system to maintain reliability for known users while extending adaptability to unknown users through peer-to-peer relay.

The patent implements self-service by allowing devices to automatically broadcast and relay voice commands without requiring manual user registration or configuration. When a user speaks to a device, the system automatically detects whether the user is known, broadcasts the command to appropriate devices if necessary, and receives execution results without user intervention. This eliminates the need for users to manually register or configure their profiles while maintaining precise user identification through automated voice recognition.

The patent applies partial action by having devices broadcast only when necessary (when the user is unknown to the local system) and only to relevant devices in the network. The broadcasting mechanism includes filtering and selection logic that determines which devices should receive the voice command based on user recognition capabilities. This partial broadcasting approach extends adaptability to handle unknown users while avoiding the excessive complexity of implementing universal broadcasting to all possible devices.