I've been around many technologies and love IT. Although I dealt with speech-recognition a lot (including AI and NLP), that doesn't mean I am cornered to only doing that. Producing a strong speech-recognition process means producing a strong and reliable back-end, and that is needed a lot for other purposes also.