The Mechanical Turk was a chess playing curio of the 1700s. This automaton could beat the greatest chess players of the courts of Europe, exhibiting astonishingly lifelike behavior. At the time, none could discover its secret…a little person hiding inside the contraption, operating the controls.
Thus the Mechanical Turk was an early example of “artificial-artificial intelligence.”
Enter Amazon’s Mechanical Turk service of today. This allows users to submit a task which can be performed by a human with relative ease but which still challenges AI, such as image or affect classification. Tens, hundreds, or thousands of workers perform this task in parallel, thus returning the results very quickly to the user.
How does Amazon’s mechanical turk parallel the original? How does this story inform our understanding of AI today?