// TALK · DEED #4 · 13 JUNE 2017 · WITH GIULIO CALVANI
Parole, parole, parole: voice interfaces need dialogue
Given at DEED #4, an AperiTech evening on vocal user interfaces, with Giulio Calvani, screenwriter for I Cesaroni and Il commissario Rex and my partner in The Land. We started from the 2017 status quo, Apple HomePod, Amazon Echo, Google Home and Microsoft Cortana, and from the feeling that talking to machines is a déjà vu.
Voice interaction comes in four modes with different results: command, dictation, agent and identification of the speaker. Ben Sauer's line frames the design problem: the medium you design in is not the medium you experience, because you draw logic diagrams and never hear the script. So who should design these dialogues? Our answer was screenwriters and developers in the same room, since a dialogue that works needs characters and clashing points of view, and those are a writer's tools. 28 slides, about 500 views on Slideshare.
Voice interaction comes in four modes with different results: command, dictation, agent and identification of the speaker. Ben Sauer's line frames the design problem: the medium you design in is not the medium you experience, because you draw logic diagrams and never hear the script. So who should design these dialogues? Our answer was screenwriters and developers in the same room, since a dialogue that works needs characters and clashing points of view, and those are a writer's tools. 28 slides, about 500 views on Slideshare.