H.I.T.

In a performance following a known format of game shows two teams with 3 members compete in developing an image recognition algorithm. The three rounds of the game show presents various stages of developing an object recognition model that requires human labour in form of so called “Human Intelligence Tasks”(HIT). The two first rounds are often outsourced labour performed for free  by social media platform users or click workers earning a little money for “easy tasks” on platforms such as Amazons Mechanical Turk. The third round is performed by an imagined developer. During the first round competitors are labelling images either as Labour or Not Labour. In the second round the task is to describe the images. In the last round the “developer” categorising the images into “automated”, “unskilled”, and “skilled” labour. In between the rounds the audience are presented with videos describing aspects of the labour that the contesters are performing. Also “the professor” -a voice from above are sharing his rather cryptic knowledge of what is happening on the stage. In the end the developed image recognition algorithms are tested in a “face off”. The algorithm of each team gives a result with a confidence level expressed in % classifying the image to represent  “automated”, “unskilled”, or “skilled” labour. “The professor” is playing the role of the jugged providing the ground truth that the result are compared to. Team B is declared the winner and are presented with a H.I.T check of 4,92€ symbolising an estimated sum of money earned from the labour performed by the team members on stage that night.