There I was, two weeks ago sitting in an auditorium on a Sunday attending TEDxCMU. It was an inspiring and emotional day with twelve fantastic talks from local and national presenters.
One of the talks was by Luis von Ahn, a CMU professor. It only took me a minute to realize that I was in the presence of an amazing innovator, and that I had the privilege of being one of the first people to hear about his latest project.
All the confirmation I needed that what I witnessed was important came nine days later when TechCrunch went public with what it knew about his project: Meet Duolingo, Google’s Next Acquisition Target; Learn a Language, Help the Web.
You see, von Ahn had previously developed a crowdsourcing project for labeling images, which Google bought. Then he came up with reCAPTCHA, where the action of completing a CAPTCHA to prove you were human was actually simultaneously helping to digitize books (the letters and words you identified were actually scanned from books). Yup, Google bought that too.
Now with Duolingo, he has a modest goal: to translate the entire web into every language, for free. What’s so innovative about this is not so much the technology underlying it (which I don’t think anyone including me has seen yet), but instead the incredibly simple but disruptive business model. You see, on the one hand people pay billions of dollars per year to learn a new language. Rosetta Stone itself is a $250 million/year business. On the other hand, it would take many, many billions of dollars to pay people to translate all of the web pages in existence into other languages. Duolingo aims to provide users with a free way to learn a language while harnessing their learning activities to translate the web.
It is worth pondering the question, for just a minute, of whether this is a net “good” thing. Our economy is based on monetary transactions – I do something for you and you pay me, then someone does something for me and I pay them. Rinse, repeat. Duolingo is obviously non-monetary. My first instinct was to call it a barter transaction, but it really isn’t that either. Duolingo is something else – its an “efficiency play”. Duolingo captures current “waste” (your language learning activities that before were just helping you learn a language), turns that waste into “value” (translating web pages, which a company might have been previously willing to pay people for), and uses that newly created value to reduce your language learning costs to zero.
The language learners win, they get to keep their money. The companies willing to pay for web page translations win – even if Duolingo charges them, it will likely be a fraction of what they would have paid. People everywhere also benefit, with web pages more accessible to readers around the world.
The losers are anyone that makes money teaching a language (Rosetta Stone), and people that make money translating. It seems to me though that the efficiency created by Duolingo is not just a redistribution of money between parties, but a net reduction of monetary transactions. It reminds me of the oft-quoted statistic that Craigslist turned the $100 billion dollar classifieds industry into the $100 million dollar classifieds industry.
It should be pointed out that paying for translation is expensive, and that Duolingo will result in the translation of web pages that would not otherwise be translated. In other words, it won’t take money away from existing translators in a lot of cases, because most of the web pages would never have been translated anyway. So in those cases, it is a clear positive outcome.
Still, it has me wondering about the winners and losers of this game.
Rocco Tarasi was an accountant, investment banker, and CFO before becoming a technology entrepreneur.