| Someone should make an app for this and it could be done way better. Here are is how I would do it so its better and fits in the use case. 1. More realistic and even custom sounding text-to-voice API such as amazon polly or - https://aws.amazon.com/polly/ (this one uses polly https://www.text2voice.org/) or google wave net https://cloud.google.com/text-to-speech/pricing 2. Then, and this is where you can get creative, is to mask the app so its UI looks like something that is allowed in a classroom i.e. dictionary, wikipedia, ebook reading, audible, spreadsheets etc. Also, make sure to name your app something innocuous. See these examples of Reddit masked as Microsoft Outlook:
http://pcottle.github.io/MSOutlookit/
Reddit as word -> http://pcottle.github.io/MSWorddit/
Reddit as code -> http://codereddit.com/
Reddit as unix shell -> https://redditshell.com/ Also, allowing the app to recognize which companion you are voice to texting (with the earpods still paired with original phone), then the app can use logic to know when one person is "speaking" and hold off on speaking a message until your companions message is over, otherwise there is potential for overlapping cross-talk (also have quick-cancel/pause speaking). Other obvious features would be to clear spoken texts from the input box and maybe even keep recently typed/spoken words as in a "recently used" area to minimize new full-text input being required. And if partner's app is doing the text-to-voice, it would already have the conversational nouns and details. Anything to minimize typing input, etc. If someone builds this lmk :) |