An amazing interview at Spark with Luis von Ahn, inventor of those blurry text boxes that you have to type in to prove you’re a human when you want to fill out a form. Ahn is one of those people whose mind just solves things. So after feeling guilty that he was making human beings type in so many extra letters in their lives (several billion a month), he decided to figure out if that wasteful activity could be made productive. He created Recaptcha, where the words you type came from public digitization projects he was working on where the OCR process had failed. In other words, he was running words by people to interpret them to aid computers in solving scanning problems. It’s like the digital equivalent of the farmer, for whom no materials go to waste. If someone is already doing something, so Ahn, how can that aid the larger public good?
Now he’s creating a project called Duolingo, and his aim is to translate the entire web. Yes, he’s ambitious. How do you get people to do this who have no knowledge of a foreign language and do it for free? Sounds impossible? Not for Ahn. He’s created a language learning company, Duolingo, that is expressly based on learning through translation. The sentences you learn from are drawn from websites that need translating. As you learn a language, you are translating the web. There are then voting systems (which translation do you think is best?) to help refine the process. And there you have it — learning and work on a massive scale.
Some will say that teaching people a foreign language via translation is pedagogically suspect. Not so. The history of language learning has traditionally utilized translation as its best practice. It’s only recently we’ve switched to so-called “communicative” models — with very mixed results. Check the historical record for people’s ability to read multiple languages and you’ll see what I mean. Some might also object that there is no way these “learners” can translate as well as experts. Again, not so. According to Ahn, he is getting results that match expert test groups. Given a prescribed task, the crowd in this case is as smart as the experts.
The bigger point in all this, however, seems to be the way Ahn is a master of theorizing work as an unconscious process, something we do without knowing it, something we do while we do something else. There is a bit of Facebook theory in this: we are all working for Facebook everytime we add something to their site which allows them to monetize our statements of self-representations, but we don’t think of this as “work.” Ahn’s model is similar but instead of getting something amorphous — social capital — in order to get something amorphous — lexical identities that can then be the object of advertisements — he’s interested in giving you something concrete (a new language!) in return for your help in doing something concrete, in this case, something socially beneficial, like making the web available in multiple languages.
Ahn has brilliantly combined labor and the unconscious. It’s like Marx and Freud have come back in the same figure. Now where is Nietzsche…