The company, which claims it has invested over one million research hours on speech recognition during the last 20 years, has developed newer technologies that will not only make traditional voice-activated services more accurate, but will extend the use of voice input to a host of other communications technologies.

Most of the technologies demonstrated by AT&T Labs this week still have a long way to go before they hit the commercial marketplace. Researchers don't yet know when – or if – many of the technologies will find their way into real-world applications. Some technologies developed by AT&T recently, however, have already found their way into products offered by both the telecom giant and its partners.

For starters, AT&T has developed a core technology platform called Watson that is a cloud-based system of services that identifies words and interprets the meaning and context of those words to deliver more accurate results. The system is built on servers that model and compare speech to recorded voice. Watson is a “smart” platform: with more data, it is able to adapt and learn so it can improve accuracy. It can also cross-reference data to use speech as input for a variety of communications.

“We are really on the cusp of a technology revolution in speech and language technology,” said Mazin Gilbert, executive director of speech and language technology at AT&T Labs. “It's no longer about simply trying to get the words right. It's about adding intelligence to interpret what is being said and then using that to apply to other modes of communication, such as text or video.”

Some of AT&T's research has borne fruit recently. At a recent demonstration of the company's more cutting-edge applications, the company showed off its iRemote application, which turns an Apple (News - Alert) iPhone (and some other smartphone devices) into a voice-activated television remote. The application allows users to speak normal sentences to initiate a search for specific shows, actors or genres.