Occasionally, we come across geniuses who speak multiple languages, sometimes 7 or 8 even, and we are like, “wow how do they do it?” Some people are naturally inclined to pick up a language while others find greater comfort in numbers. Be that as it may, around the globe, there are approximately 7k languages spoken, of which 3k are at serious risk of disappearing. When a language disappears, the loss is as great as an endangered species withering away. Something (a cultural aspect native to few people) will forever be lost and humans will become poorer. Languages that aren’t used in business or the pursuit of science, are at greater risk.

That’s why we have to find a way to preserve them.

The time-honored tradition of storytelling from which the modern business approach is also shaped finds its roots in the oral form of communication, particularly in native languages. Thankfully, Google has paid heed to the warnings of linguists and anthropologists and has taken a major step forward in preserving some of the endangered languages.   

Woolaroo was launched and it supports 10 languages - Calabrian Greek, Louisiana Creole, Māori, Nawat, Tamazight, Rapa Nui, Sicilian, Yang Zhuang, Yiddish, and Yugambeh. Their Innovation Partners and Arts & Culture group conducted a survey to find out the lesser-known languages that employees speak. They worked with these individuals to come up with a dictionary in the respective language (continues to be WIP). Compiling the English dictionary required a Johnsonian effort (Re: Dr. Samuel Johnson) which was actually a Herculean task. Sorry, there was no “Dr.” Hercules – unless he did the disappearing act as well.  

Google Cloud offers two computer vision products – AutoML & Vision API.

AutoML: “Simply upload images and train custom image models with AutoML’s easy-to-use graphical interface; optimize your models for accuracy, latency, and size; and export them to your application in the cloud, or an array of devices at the edge.”

Vision API: “Google Cloud’s Vision API offers powerful pre-trained machine learning models through REST and RPC APIs. Assign labels to images and quickly classify them into millions of predefined categories.”

Woolaroo is an open-source photo-translation platform powered by machine learning and image recognition. Built on Google Cloud, users can explore endangered languages around the world. Real-time pictures taken will be returned with translation in the selected language along with their pronunciation.

Some of these languages – for instance, Yugambeh, a native Australian language – may not have the exact words for modern-day indoor items and the Open API interface allows the dictionary to be updated based on inputs. Users can add and edit and make a significant contribution to the repository of existing knowledge. Yugambeh has about 100 speakers and is at serious risk of being forgotten. The way to preserve such languages is through a participative approach and technology doubling up as a global unifier.

Google Arts & Culture initiative is driving this and it should pick up in a big way. It should also encourage other users to take a shot at learning these languages. There are 3000 languages that are at risk and there’s a massive amount of work to be done. I am certain there will be many languages that are spoken today by fewer than 30 or 40 people. 

It’s mind-boggling to imagine what computers can do was endeavoured by humans hundreds of years back with pen and paper. The dictionary that we take for granted is a result of thousands of millions of man-hours. Languages develop over many centuries so let’s do all that we can to preserve them