Joe Levi:
a cross-discipline, multi-dimensional problem solver who thinks outside the box – but within reality™

Prediction: Voice Recognition and Text-to-Speech

Voice Recognition is getting pretty good these days… so is Text-to-Speech… but what would happen if you combined them?

Imagine a world where you could have 95% accuracy in voice recognition, and 95% intelligible text-to-speech. We’re pretty close to that now.

Add one other component: I’ll call it Voice Font(tm). Imagine a text-to-speech “font” of your own voice. Where anything you type can be run through a font and made to sound as if you’d said it, just like a Font applied to plain text might make something appear in calligraphy.

Tie it all together: Talk into your phone, your phone converts your voice into text then sends your Voice Font and message to the recipient. The recipient’s phone (computer, whatever) then receives text and converts it into YOUR voice. They respond in kind.

The potential bandwidth savings by sending text only is staggering. We’ve been focusing on how to expand our bandwidth, we can just optimize our bandwith usage.

Share

You may also like...

Leave a Reply