An English In, Japanese Out Computer

Computers that can understand what people say, then instantly translate the spoken words from one language to another, may soon arrive--finally. For years, researchers have been inching toward systems that combine "voice-independence" with natural-speech processing. Today, you can have one or the other, not both. For example, computers that can comprehend strings of naturally spoken words must be trained to become familiar with the unique vocal pattern of each user.

To continue reading this article you must be a Bloomberg Professional Service Subscriber.