*chuckle*
A little trivia for everyone: Did you know that IBM has been working on that for more than FORTY (yes, 40) years?!
"May be difficult part" is really an understatement. 6 billion people on earth. Hundreds of thousands of dialects. Accents. Voice coaching, speech therapy, and makes speech RECOGNICION nigh impossible in real time.
But, of course, Shiva was making a point that "text to voice" may be difficult, not "voice to text" which is nigh impossible. Actually, text to voice is possible, now, but it's still at the monotone, robotic stage. At least live humans voicing for you put some inflection into their voice.
The technology you want is still many years off. Do you want to wait until the "simple solution" comes out in a couple of decades, or are you content to use relay until the right stuff comes along?
