While there are more than 7,000 languages spoken worldwide, only a select few are well-represented across online spaces.
To address this issue, Google has introduced the 1,000 Languages Initiative – a commitment to creating an AI model capable of supporting the 1,000 languages spoken most often around the world.
Google believes this initiative will help to create a more inclusive online environment for billions of people in marginalized environments.
Notably, Google acknowledges this initiative will take some time. However, the company is already making meaningful strides toward its goal.
As this technology continues to evolve, changing the way people live and work, Google believes it’s crucial to create an ecosystem where all languages are properly supported in the digital world.
From a CX perspective, the announcement is significant. With advanced automatic translation capabilities through Google, companies may one day harness this AI model to serve up more relevant experiences to customers across the globe.
From 400 to 1,000 Languages
As part of Google’s new language-focused initiative, and its growing focus on multi-modality, the company has created a Universal Speech Model (USM).
The model is trained in over 400 languages already, giving it the most coverage of any speech model to date.
To expand on this work even further, Google has committed to partnering with communities around the world to source representative speech data.
The organization also recently announced new voice typing options for nine more African languages with Gboard, thanks to its work with researchers and organizations within the African landscape.
In South Asia, Google is also actively working alongside NGOs, local governments, and academic institutions to collect representative audio samples from across all regional dialects and languages.
With such thorough research, it’s fascinating to consider how the upgraded USM may enhance Google’s latest voice innovations.
Indeed, as Calum Barnes, Head of Product for Cloud Speech at Google Cloud, stated during the recent Google Next event:
Our goal at Google Cloud is to take all of this technology, and all the expertise around speech and voice built up at Google over the years, and make them available for enterprises to use in their own applications and to power their own voice experiences.
At Google Next, the technology pioneer also announced the general availability of its CCaaS platform as its enterprise business begins to ramp up.