$15/hour contractors fuelling ChatGPT's power
Tech empires have been built with the labour of lower-skilled, lower-paid workers throughout history
OpenAI and ChatGPT rely on a team of contractors to perform essential tasks in the development of AI systems. These workers are responsible for training AI models to analyze data and produce convincing text and images, which are crucial to cutting-edge products like ChatGPT. Despite their crucial contribution, these contractors are paid starting at just $15 per hour, without any benefits.
One of these workers is Alexej Savreux, a 34-year-old who has worked in a variety of fields but currently works as an AI trainer. Savreux labels photos and makes predictions for text generation in applications, among other tasks, to improve the accuracy and effectiveness of AI. He acknowledges the importance of these contractors, stating, "We are grunt workers, but there would be no AI language systems without it. You can design all the neural networks you want, and you can get all the researchers involved you want, but without labellers, you have no ChatGPT. You have nothing."
While the work of contractors like Savreux is essential to the success of AI technology, they often go unrecognized and undervalued. According to Sonam Jindal, program lead for AI, labor, and the economy at the Partnership on AI, a San Francisco-based nonprofit focused on AI research and education, the AI industry depends heavily on a large human workforce. However, these workers are often anonymous, with credit primarily attributed to tech startup executives and researchers.
Unfortunately, this problem is not unique to the AI industry, as tech empires have historically relied on the efforts of thousands of lower-skilled and lower-paid workers to build their computer empires. From punch-card operators in the 1950s to recent Google contractors who have expressed concerns about their second-class status, the industry's dependence on contract workers is not a new phenomenon.
The AI industry is following a similar trend, embracing an unsteady, on-demand nature of work. Individuals are employed through written contracts either directly by companies or via specialized third-party vendors focusing on temporary work or outsourcing. Health insurance and other benefits are rare or nonexistent, reducing costs for tech companies. In 2021, the Partnership on AI warned of an imminent surge in demand for "data enrichment work" and recommended fair compensation and improved practices. The organization also published voluntary guidelines for companies to adopt, although only DeepMind, an AI subsidiary of Google, has publicly committed to following them.
The recent formation of a union by AI contract workers in Nairobi, Kenya, who have worked for Facebook, TikTok, and ChatGPT, highlights the low pay and mental toll that the work can have on these employees. As demand for this type of work continues to surge, it is crucial for companies to recognize and respect the contributions of these workers. According to Jindal, "This is a new job that's being created by AI. We have the potential for this to be a high-quality job and for workers who are doing this work to be respected . Despite the recognition of the significant role of AI contract workers, they continue to receive low wages and minimal benefits. This issue is not limited to the AI industry, but it is particularly alarming given the projected growth of AI. According to a report by the McKinsey Global Institute, AI could potentially generate $13 trillion in global economic activity by 2030. The benefits of AI will be enjoyed by many, but it is vital to ensure that the workers who fuel its development are not left behind.
In conclusion, AI contract workers like Alexej Savreux are the unsung heroes of the AI industry, responsible for the significant progress and achievements of AI systems. Their contributions are crucial to the continued growth and success of AI, and it is essential to recognize and value their work by offering them fair compensation and benefits. As we continue to rely on AI for various aspects of our lives, we must also prioritize the welfare of the workers who make its development possible.