Author: Lilly Irani
Publisher: Public Books
Publication Year: 2015
Summary: The following article describes how when things in the technology world become automated, the work that is being “replaced” is not actually replaced, but displaced. For example, a manufacturing process that has been automated may replace the individual workers, but over the long term, it just displaces them to monitor the machines and facilitate the automation. Simple tasks that require human attention can be outsourced through services like Amazon’s Mechanical Turk, such as image recognition for a machine learning (ML) training set or basic data cleaning. These crowdsourced workers rarely make minimum wage and are often hidden by the companies, usually tech, whose algorithms they enable. It is important for data scientists to understand where data comes from, what labor and processes have gone into producing and cleaning it, and have ethical consideration for this work, especially in huge artificial intelligence and ML models.