Automated call centers are only the most obvious way speech recognition will be used. The software is now becoming sophisticated enough to identify speakers through 'voiceprints,' akin to fingerprints, eventually reducing the need for personal identification numbers.