Technical note on the terminology here. Just FYI, this is not strictly speaking crowdsourcing. Just like outsourcing refers to replacing local workers with workers in a foreign country, crowdsourcing refers to replacing people who have a job with generic members of the public. Amazon's Mechanical Turk is crowdsourcing.
von Ahn's area of work is in human computation, which specifically refers to the augmentation or replacement of some part of a computer-based computation process with human calculation. CAPTCHA is a good example of human computation, because it augments the problem of OCR by sourcing the identification of OCR-hard characters to people on a massive and automatic basis. Human computation allows you to formulate algorithms that are too hard for computers to do and too laborious for people to do into a sort of "average" between the two. This gives you a lot of leverage and allows you to do some neat stuff, like the above.
If you didn't know, von Ahn's main claim to fame is having founded re-CAPTCHA and then having marshaled the company through a successful Google acquisition. After that I think he became professor at CMU, where he is currently.
EDIT: a more scholarly examination of the differences between human computation and related fields is here:
I'm pretty sure he was already a professor at CMU when he created and sold reCaptcha. Also of note, he won a MacArthur Genius award shortly after reCaptcha became mainstream.
This is pretty awesome. I am very curious about language learning especially if it can be made more addictive(I have started playing with it myself http://thedjpetersen.github.com/playground/language_game/). I especially love the concept they are presenting I would like to see how they design the interface. How easy it is to go from sentence to sentence.
I have always thought that if you could simplify language learning enough and maybe add some of the addictive game elements that games like World of Warcraft have, people would not look at it as a chore but rather pour themselves into to it.
Yea, that was a great talk - I was there. I love the n-sided task arbitrages this guys comes up with. Is there a word for these? I mean "games for good" is a subset, crowdsourcing a superset, but is there anything more specific?
von Ahn's area of work is in human computation, which specifically refers to the augmentation or replacement of some part of a computer-based computation process with human calculation. CAPTCHA is a good example of human computation, because it augments the problem of OCR by sourcing the identification of OCR-hard characters to people on a massive and automatic basis. Human computation allows you to formulate algorithms that are too hard for computers to do and too laborious for people to do into a sort of "average" between the two. This gives you a lot of leverage and allows you to do some neat stuff, like the above.
If you didn't know, von Ahn's main claim to fame is having founded re-CAPTCHA and then having marshaled the company through a successful Google acquisition. After that I think he became professor at CMU, where he is currently.
EDIT: a more scholarly examination of the differences between human computation and related fields is here:
http://alexquinn.org/papers/Human%20Computation,%20A%20Surve...