Google is updating the Google Assistant with a host of new features to make interacting with it a more natural experience. This comes not only as a result of adding new voice options, but also increasing the conversational capabilities of the AI assistant.
Firstly, there are six new voices being added to the Google Assistant. These have been developed with the addition of a new technology known as WaveNet, which uses machine learning to affect the cadence of the Assistant’s voice. Making it sound more natural and human-like.
This update will also see the introduction of Google Assistant’s first celebrity voice in the form of singer John Legend. Which is apparently coming after the other new voices have been implemented.
Also making things more natural while interacting with the Google Assistant is a new recognition system that removes the need to keep saying “Okay Google” before each request. The AI will be able to recognise instructions and input after the initial activation phrase; and will also realise when the user is no longer interacting with it.
Called Continued Conversation, it will be added to the Google Assistant for users in the coming months. At the same time, a Multiple Action feature will be added. This simply allows users to make multiple instructions during the same sentence. Something that sounds simple for humans, but apparently very difficult for machines to understand.
Finally, Google is addressing parental concerns that the AI assistant is teaching kids to be rude. A new optional ‘Pretty Please’ option will be added later, which will thank users for asking the Google Assistant to do things nicely. Namely, by saying “please” and “thank you” when making a request.