In recent days I’ve used a lot of Amazon Polly and Mac’s say command to generate speech. I still think we have a lot of unlocked potential in user interfaces utilizing speech, especially I think of systems that talk to you first and you respond.