Abstract :
The future demand for speech and language technologies looks promising, but the difficulty of creating robust, high performance components delays the day when effective applications can be deployed. In the meantime the possibility increases for those applications to be satisfied by other technologies (for example video phones over high bandwidth phone links to mobile hand held devices). Unless speech and language technologies can be used to create robust, high performance components within reasonable time frames, the current boom in speech and language applications looks unsustainable except in specific niches such as telephone dialogues, and database (internet) information access
