Her Voice

 

As I've been exploring the text-to-song idea I realize that it's still in its talking-head phase, in the transition between the spoken voice and singing voice. Eventually, text-to-speech voices will be text-to-song, with a melodic prosody to them. It's kind of an exciting idea, and will probably come very quickly because we're already at the stage where it has become an inflection point.  In the early days of the internet, streaming video was the big wish. They were very small and took a long time to download. It took about 10 years for that to completely mature. With the exponential pace of technology, those talking-head phases are shorter.

1/7/2024

5/16/2024: The use of  the "Her" (Scarlett Johanssen) voice in ChatGPT-40 is an interesting development, but it's still another example of a talking head.