Dubzer, Microsoft and Google are asking the same question. How will translation become ubiquitous? (Part 1)

by Santosh

Thinking about how much content is created every day on the Web just makes our heads spin. Just getting a small portion of this content to another language ought to be valuable for some users. There is evidence to suggest that this might be true. However, evidence also tells us that the web is just as local as it is global.

We’re not the only ones who have been thinking about this problem. Dubzer was originally inspired by an old interview of Google co-founder Larry Page talking to a classroom of graduate students. Larry was very candid. He believed that there was a lot of information that is not available in the user’s language. He also said that the problem remains unsolved by purely machines.  The human part of translation cannot be ignored and we got thinking – may be a startup can approach it from the human side and find a solution. That thought in 2009 inspired the creation of Dubzer.

Along with Dubzer, Microsoft Research and Google have attempted various solutions to the same problem. They have learned similar lessons as we have, that mirroring content to create copies in another language alone is not enough.

Well, at least no one can fault us for not thinking big!

So what makes us different? In many ways we do operate a lot like a research lab. We embrace failure and focus on maximizing our understanding of how natural languages and the web interact. That is probably where the similarity ends.

We firmly believe in the power of natural selection and the ability of markets to self-select solutions. This is why we decided from day one that Dubzer was to be a product startup. If Dubzer cannot create sustainability, it probably does not deserve to be around. After all, a problem delayed is a problem denied.

We believe that content-owners, application developers are not going to solve this problem by themselves. For them, the economics of internationalizing content is a significant barrier. This is where we depart significantly from Google’s and Microsoft’s approach to influencing the Wikipedia / Wikimedia sphere to translate itself.

To help understand this better, have a look at how Facebook, Twitter, TED and Mozilla Firefox have managed to internationalize themselves. Since they lie in the intersection of popularity on the web and an international audience, they have been extremely successful in recruiting bilingual users and translators to help them cross over the language barrier. After all, a small percentage of several million is a significant number to create value through translation.

What we gather from these examples and from our own iterations/data is that given the right tools and the right incentives, online users are able to successfully surmount the language barrier.

How do we make this force available to the several million content creators who are online today? The key to solving this larger puzzle is to open up the constraints and connect the incentives. Over the next few months, we intend to answer this very question.

“If language had been the creation not of poetry but of logic, we should only have one.” – Friedrich Hebbel.