What Happened
Workers are manually labeling thousands of photos and videos to train artificial intelligence systems. They identify and tag everyday objects like cars, dogs, traffic lights, and furniture. This labeled data becomes the foundation for AI systems to recognize and understand visual information.
Why You Should Care
Every time your phone unlocks with face recognition, your car's backup camera spots a pedestrian, or Instagram suggests tagging your friend, it's because someone got paid $15/hour to click on a million eyeballs.
📚 The Basics
AI doesn't naturally know what anything is — it has to be taught like a toddler, but with math. Machine learning works by showing computers millions of examples: 'this is a cat, this is a dog, this is a cat wearing a hat.' The more labeled examples, the better the AI gets at recognizing new images. This process is called 'supervised learning' because humans supervise the AI's education by providing the right answers.
🧠 Look Smart At Dinner
Say This
The dirty secret is that 'artificial intelligence' is really just thousands of people in developing countries clicking on pictures for pennies.
Context
Companies like Scale AI and Labelbox employ hundreds of thousands of workers worldwide, mostly in Kenya, Venezuela, and the Philippines, to create training data.
Avoid Saying
Don't say 'AI will replace all human jobs' — right now AI literally depends on human jobs to function at all.
The Approved Opinion™
“It's important that AI development creates meaningful employment opportunities while ensuring fair wages and working conditions for data workers.”
