By Stephen Schenck | April 19, 2012 1:26 PM
Siri on the iPhone 4S did a lot to raise the profile of speech recognition on mobile devices. While Siri’s real accomplishment is in its ability to parse language, the speech-to-text itself is a vital component. Of course, Apple’s not the only one around with an app that accepts voice commands, and there are various solutions available on all the major platforms. With all that language-processing tech available, do we need another option? AT&T seems to think so, and today announced the future availability of its Watson system to developers.
This June, AT&T will release APIs for developers to incorporate the company’s Watson speech recognition system into their own apps. It didn’t discuss platforms, but considering its Watson demo app is available for Android and iOS, we’d say those are safe bets.
AT&T’s goals for Watson are admirable, offering developers customized APIs for maximizing performance based on how users will interact with an app; APIs for web searches, text messaging, and getting information on local businesses will be among those first released. That might attract some interest, but are developers really interested in moving away from established speech recognition systems? Android, for example, already provides hooks for such services, and taking advantage of Google Voice Search might make more sense for many developers than taking a chance with AT&T and Watson. By the time summer rolls around, we should start to get a better picture if Watson will be the success AT&T is hoping for.