Currently the clod sound setup assembles spoken phrases from individual recordings of words. That's why a sentence from a British pilot for example will be composed of a variety of intonations at once. The words sound as if they are being pulled out of a hat and range in inflections varying from mild surprise, grim determination, cheerful optimism and mincing queer, all in the one sentence. If they were to try and organize something as 'complex' as a shout of alarm, they would need to revise their entire system of AI speech routines.
|