This is great, but there remain very difficult problems to be solved. The prosody generated by this is fairly generic and not informed by a true understanding o

This is great, but there remain very difficult problems to be solved. The prosody generated by this is fairly generic and not informed by a true understanding of the text. Consider this sentence: I have plans to leave. If you stress the word "plans", the sentence means that the speaker is not necessarily intending to actually leave. However, when the stress is on "leave", the speaker definitely intends to leave. A human reader can easily infer the correct meaning from context but text-to-speech systems can't because they don't have any systematic understanding of the things being talked about and the social pragmatics of the discourse. As long as these issues aren't solved, text-to-speech systems will make mistakes. These mistakes will be easy to spot in some cases but can also have catastrophic consequences in other cases: "I have plans to bomb North Korea."