Serious question, how does this stand out from speech synthesis we have had for years? The only problem that I see that this solves is the different pronunciations of "read".