When voice isn't working because the JSON model is invalid or missing, we should throw an error. This will help prove and debug issues where no voice instructions are coming through regardless of mute state. It's especially important because we're attempting to support a variety of backend APIs.
When voice isn't working because the JSON model is invalid or missing, we should throw an error. This will help prove and debug issues where no voice instructions are coming through regardless of mute state. It's especially important because we're attempting to support a variety of backend APIs.