Artificial Intelligence has been the most sought after technology for home sound and assistance. For a long time, AI has since been in the hands of the giant technology companies like Facebook, Amazon, Apple, and Google.
With AI advancements and more skills added every time, these companies have a large source hub for collecting data they need so as to make it even smarter.
Mozilla is now joining the race of AI voice-recognition software. For AI to be more accurate, it needs to gather several data and voice samples in its intelligent domain in order to respond to human voice interaction in a smarter and accurate way. This has been possible with companies like Google which has the already established source of data needed to get started.
Making it Open Source
Mozilla might get a hard way to make this race successful. As it begins its journey, the Mozilla Organization is asking the public to contribute their voice samples and other important data in order to get started. Yes, it has to give the power to the public by asking users to pool information they need. However, to achieve all this, they have to make it be an open source AI.
By using the Common Voice project, users will be able to donate their voice samples in a similar system to how Alexa and Siri work.
Speaking with The Verge, Sean white, who is the vice president of emerging technology at Mozilla says that “”Currently, the power to control speech recognition could end up in just a few hands, and we didn’t want to see that,”
Those who are interested in contributing to the Mozilla open source AI software, they have to head over to the Common Voice in order to take part in reading out some sample sentences. Those willing to share their age, gender, location, and accent will have help the AI reduce the error rate of recognizing various accents.
“For us to be successful with data commons, there has to be a motivation [for users] other than realizing one day that they’ve been giving away all their personal data…We have to make their experience better because they’ve participated.” says White.
Yes…But it takes time…
This task may take a long time, their focus is to collect as much accent as they can so as to make the open source AI even smarter.
As the race starts, they need a lot of help even from third party sources for some input. Relying on open source alone may not make them beat Google but will be of great benefit to the voice recognition Software. Chris Nicholson had to say that “We may need third parties to step in – NGOs, governments, coalitions of smaller private firms – and pool their data.”
We eagerly anticipate to see the next open source voice assistant rolled to the public. Somehow, this may take some time but it will be worth helpful.