The Alexa team at Amazon have been working on modifications to the Speech Synthesis Markup Language for the US, the UK and Germany to give developers tools to take the voice assistant from literal dictation to humanized being.
New SSML tags will allow Alexa to whisper, censor expletives, put emphasis to her read or change the entire prosody of her speech (rate, volume and pitch). Another tag will enable Alexa to read substitute text in place of what might be displayed.
Along with some humorous purposes, developers could theoretically have Alexa whisper one-time passcodes or blank out keywords in quiz games. She could even become a decent bedtime storyteller.
That’s the sound of progress for many apps and many businesses.