Voice synthesis algorithms

What's the latest state of the art for voice synthesis? Is there real-time neural network models for audio synthesis yet? I wish to roleplay as a 2hu on vrchat and not sound like a robot. Is this going to be possible in the near future?

https://my.mixtape.moe/nzolkc.ogg

>>1
vocaloids

>>2
nostalgia
https://www.youtube.com/watch?v=kvnIFo3xMfY
I can only imagine what the first meeting for vocaloid development was like
``hey, what if microsoft sam was cute?''

Synthesizing speech is impossible without Strong AI understanding what it reads and emulating emotions. Without emotions speech sounds dead.

>>4
Moreover, AI needs personality, as the speech of a Russian drunkard would be different from speech of a Japanese girl.

>>5
I.e. an angry drunkard is incomparable to an angry schoolgirl.

>>4

Without emotions speech sounds dead.

just like me