Reviews coming back from papers and grant proposals of mine that use (or propose to use) Mechanical Turk have included a common sentiment: Neat ideas, but what happens if Mechanical Turk goes away or if wages don't stay as cheap?
Mechanical Turk is an interesting and available vehicle for research in human computation, but it's not actually necessary. It's also interesting but not necessary that the wages are so low currently. What is important is the new models of interaction that we're exploring with human computation, and the new models of integrating human and machine work that platforms like Mechanical Turk enable.
Leave a comment