|
In terms of voice/speech, we maybe talking on different points. I'm considering voice as an input mechanism, but the output mechanism certainly has to be dynamic and use the most sensible mechanism in the given context. Ex: 1. Alexa play song --> output on in-built speaker 2. "Alexa turn on Warriors game", "Alexa play latest episode of Game of Thrones" --> output onto nearest TV 3. Alexa get direction to San Francisco --> sends to my phone screen 4. Alexa show me top Sushi restaurants nearby --> send to nearest display ( TV or phone ) ... and so on. So yes, I definitely agree with you, voice as an output quickly becomes untenable. But again, that's what I mean by ubiquitous, you're no longer tied to a device for input/output. Your environment/context will define what your input/output mechanisms are. Outputs can be any displays, speakers, TVs, thermostats, lights, etc. And in most cases, voice is the simplest, most intuitive input mechanism for simple queries that a majority of our daily interactions with our surroundings will require. Just as touch devices required UI developers to simplify their interface design to accommodate 'touch access' by removing layers of menus, pages, etc. Voice will also precipitate this type of simplification of the interface to where the core elements are accessible with simple queries, with strong, complex NLP and search behind it. And again, it's more about accessibility than it is about expressiveness through voice input. |