It’s a fact: we are not immune to the effects of the financial crisis. Faced with precarity, the misdeeds of large companies become even more palpable. In this case, not even artificial intelligence can escape it. The algorithms that power famous generative models like DALL-E 3 are trained on large datasets, many of which are classified by workers from some of the cheapest labor markets in the world.
Appen is an Australian-based platform whose job is to contribute to and improve data used in the training of artificial intelligences. Its importance is such that among its clients, we find names like Amazon, Google, Facebook, or Microsoft. The global market for data collection and categorization was valued at 2.22 billion dollars in 2022. Currently, it is estimated that this number will reach 17.1 billion dollars in 2030.
However, the high growth expectations of the industry have not translated into better wages for its workers. As reported in The Wire, Oskarina Fuentes (an Appen employee from Venezuela) explains that one and a half hours of work earns her one dollar in profits. If there are enough tasks to do, she can earn up to 280 dollars per month. To achieve this, she must be attentive to the computer for over 18 hours a day to complete as many tasks as possible.
It’s no coincidence that we are experiencing a fever for artificial intelligence. According to Florian Schmidt, author of “Digital Labour Markets in the Platform Economy,” “the industry can move very flexibly to any place where wages are lower.” In this way, locations like East Africa, India, the Philippines, or even refugee camps in Kenya are being considered as ideal places to find the cheapest labor.
Within this context, workers like Fuentes are fighting to be recognized as employees with full rights. In her words, “I would like us to be considered not only as tools that are discarded when they are no longer useful, but as human beings who contribute to their technological advancement.”
