Imagine a world where you have an artificial intelligence program on your desktop and in your mobile devices that resembles Siri or maybe even IBM's Watson. Now imagine that we don't even need type anything in order to find the information we want. In the future, the bulk of all searches could be voice activated. Some of these searches may even start circumventing Google and start going straight to the proven knowledge pools. I'm talking about sites like Wikipedia for knowledge, Quora and Yahoo answers for opinions, and other Q&A sites like stack overflow could be in the mix. This is is a major possibility that developers ought to consider.
Voice activated search that pulls from many websites should not be in the hands of only the major tech companies. Open source search technologies such as Lucene, Solr, and Sphinx ought to start gaining voice recognition facilities as well. With so few open source speech recognition technologies
out there that I can find, it seems like there needs to be more collaborative effort in this area.
Now is the time to start thinking about applications and systems that could harness this new concept of voice driven interfaces and voice activated search. We need to start solving the problems in voice recognition software for various languages and start finding work that's been shared on these fronts. The W3C is already seeing this trend and is currently investigating
how it could integrate speech recognition technologies into web standards.
The interface for search is changing. First we narrowed it down to just one unified search bar with little need to go outside of it. The next interface that will win consumers will be no interface at all. Just a microphone.
If you know of any open source projects around voice recogniton and connecting it with powerful search engines, let us hear about it. We want to take stock of what technology the general developer has to work with right now, and where we can go with it.