What are the latest advancements in text-to-voice technology, and how are they being utilized?

The training of your own voice is something that's both scary and helpful. I just hope that people will be honest about what technologies were used in the process of their recordings. I've seen video avatars with the trained AI voice of the content creator and even some of his fans coulnd't tell the difference. 
I think we need more critical approaches to what we consume online.

AI voice creation using relatively short duration recordings is getting better and better. For an artist I work regularly we created an AI voice for him, and when there are small revisions that need to be done - I've succesfully sung the parts myself and imprinted his voice over mine, so it can be seamlessly replace the part that needs to redone. No need to call the artist in. Done in 20 minutes. 

Regular speech has gotten to a point where its almost indistinguishable from real human narration. I expect a lot more audiobooks to be produced in the near future which will require almost zero editing and re-dubbing.