The article discusses possibility of speech technology as a big wave in computer interfaces. Speech technology is expected to bring the Internet to every one. The article confers to speech-recognition technology, the innovation of computing on the basis of voice properties. This technology transforms an acoustic signal into textual words and makes the user independent from constraints of general interface. The technology continues to improve but does not yet come close to approximating the speech-recognition capacity of an adult human. Although speech technology cannot be used in all situations to replace keyboards, it is already being used productively in a variety of applications and promises pervasive computing. In this regard, VoiceXML, an XML-based markup language supporting voice access of web-based services is under development at IBM Research division. There are issues in this technology, like errors due to variability of the speaker, and the acoustic environment in which the speech is produced. The key to a successful future for conversational systems is using speech effectively and unobtrusively in solutions for users' everyday problems, not just because it's cool technology.