Google has added 21 new languages (30 language varieties) to its speech-to-text recognition roster, bringing the total number of languages to 119. According to Google, using your voice to dictate a message is three times faster than typing. Based on how often autocorrect incorrectly corrects words, that's a very believable statistic.
Google's goal of adding so many languages is to make the internet more inclusive, and that's a beautiful thing. Speech recognition now supports ancient languages like Georgian, as well as two of Africa's largest languages — Swahili and Amharic. You can see the full list of newly supported languages and locales here.
The company explained that the more the machine learning models are exposed to a language, the more accurate they'll become over time.
"To incorporate 30 new language varieties, we worked with native speakers to collect speech samples, asking them to read common phrases," the company wrote in a blog post. "This process trained our machine learning models to understand the sounds and words of the new languages and to improve their accuracy when exposed to more examples over time."
Another fun feature has been added to Google's speech recognition — emoji-to-text dictation. So, now users in the U.S. can say things like "winky face emoji" or "heart emoji," and the specified emoji will appear. As of now, this feature is only available in english, but it'll roll out to other languages soon.
We kind of can't wait to accidentally overhear grown adult humans be completely serious about saying "poop emoji" into their phones.