A review and demo of the many speech and language API's offered by HTML5 in Chrome, including speech recognition, speech synthesis, named entity recognition, and translation. Minimal functional apps are given for each API.
Speech recognition is notoriously fickle and was not cooperating during my presentation. But the next day, Scott Gonzalez used it to "illustrate" his keynote speech on the fly. A brave thing to do, and the app remarkably kept up with his rapid-fire delivery. The content of his speech produced images which were only rarely appropriate. Here's the video: https://www.youtube.com/watch?v=b8fHqlz1HFI
Demo site: http://www.johndimm.com/FunWithSpeech/
Source code: https://github.com/johndimm/FunWithSpeech