Workers around the world are training artificial intelligence systems by labeling everyday objects in photos and videos. The task involves identifying items like furniture, vehicles, animals, and other common things to help AI models recognize and classify visual information.
This type of work has become a key part of AI development. Companies building computer vision systems need vast amounts of labeled data to teach algorithms what objects look like. Human workers provide that training by manually tagging images, a repetitive but essential process.
The BBC report focuses on how this labor underpins the AI industry. Workers perform the tedious annotation work that allows AI to improve its accuracy. The story highlights the human effort required behind the scenes of AI advancement, often invisible to consumers who use AI-powered tools for everything from photo search to autonomous vehicles.
The work illustrates a broader pattern in AI development. Many cutting-edge systems rely on human workers performing lower-wage labeling, annotation, and quality-control tasks that machine learning models cannot yet do independently. This creates a labor dependency that companies must manage as they scale AI products.