Portuguese 
English 
Spanish 
5 min of reading 
0 comments 
A video of a factory in India went viral on social media by showing textile workers wearing helmets with mounted cameras that record their movements. The goal is to train an artificial intelligence that, in the future, will replace these same employees. Salaries are around US$ 230 to US$ 240, about R$ 1,200 per month, in the so-called hand farms.
Social media is reacting with outrage to a video that resembles an episode of Black Mirror, but is real. The images show workers in a textile factory, supposedly in India, wearing helmets with mounted cameras that record every movement of their hands during tasks such as sewing. The stated goal is to record these repetitive movements to teach an artificial intelligence to execute them accurately, allowing humanoid robots to eventually replace the very workers who are currently filming their own routines. The paradox is cruel and became impossible to ignore when the video reached millions of views.
The phenomenon has a technical name that has not yet been translated into Portuguese: hand farms. The concept describes places where thousands of people perform simple and repetitive tasks with cameras and sensors attached to their bodies, feeding artificial intelligence systems with the necessary data for algorithms to learn to reproduce the same actions. In practice, workers receive monthly salaries ranging from US$ 230 to US$ 240, approximately R$ 1,200, to teach machines to do exactly the work for which they are being paid. Once the artificial intelligence is trained, human jobs become dispensable.
How the training of artificial intelligence by Indian workers works
According to information from the portal Metrópoles, the technology behind the viral video is more sophisticated than it seems. The helmets with cameras not only record video, they capture the perspective of the workers’ eyes in sync with the precise movement of their hands, as well as collecting data on angle, speed, torque, and other details that machine learning algorithms need to reproduce the actions with mechanical precision. Each worker effectively becomes an involuntary teacher of a student who will one day take their place.
-
Psychologists warn that today’s parents are raising an entire generation of spoiled children by avoiding setting limits, giving in to all their desires, and exchanging emotional presence for gifts and material rewards.
-
Ítalo, a young man from Pará who, even while hospitalized with a rare disease, took the Enem exam in the hospital and was accepted into medical school at three public universities: UFPA, UEPA, and UFRJ.
-
Larger than Manhattan and visible from space in a single orbital image, the Texas airport spans over 69 km² with 7 runways, dozens of kilometers of taxiways, and an annual flow of millions of passengers that makes Dallas Fort Worth one of the largest airport structures ever built on the planet.
-
He places a lemon between the positive and negative of the antenna of an AM radio tower and asks to turn on the transmitter: within seconds, the fruit starts to heat up and “sing,” demonstrating how AM radio frequency can generate vibration and sound.
The process works because artificial intelligence learns better from real examples than from artificial simulations. When thousands of workers perform the same tasks with subtle variations in technique, AI can identify common patterns and optimization strategies that engineers alone would not be able to program. Sewing in a textile factory, for example, involves decisions in microseconds about pressure, angle, fabric adjustment, and rhythm that human eyes can make without thinking, but a robot needs to learn by observing real humans doing the same thing thousands of times.
What are hand farms and why do workers accept the work
Hand farms operate in a gray area of the global labor market. For workers in India participating in these programs, the salary of R$ 1,200 per month is competitive within the local reality, and the task itself does not require special qualifications beyond performing the work they already know how to do. The agreement is seemingly simple: work as you have always worked, wear a helmet with a camera, receive your salary at the end of the month.
The problem is not the salary itself, but the long-term consequence. Every hour of movement recorded by these workers accelerates the development of technology that will make them dispensable, and they are actively contributing to their own professional obsolescence without having economic alternatives to refuse. In many cases, workers do not even know exactly what their data is being collected for, and the contracts are opaque enough that the ethical issue rarely comes to the surface. The viral video exposed what was previously happening out of the public eye.
Why Indian factories became a hub for hand farms training AI
The concentration of these practices in countries like India is not a coincidence. The combination of cheap labor, a large population willing to work in textile factories, and existing industrial infrastructure makes India an ideal environment for companies that want to train artificial intelligence with massive volumes of data collected from real humans performing manual tasks. Countries with higher wages would make the process economically unfeasible.
In addition to cost, there are also cultural and regulatory factors. Labor laws in India and other countries where hand farms operate are less restrictive regarding the use of biometric and behavioral data from workers, allowing companies to collect information that would face significant legal barriers in jurisdictions like the European Union. For corporations developing humanoid robots, this combination of low cost and few restrictions transforms developing countries into strategic hubs for training artificial intelligence.
What the video of Indian workers reveals about the future of work
The debate that the video reignited goes far beyond the momentary outrage of social media. The idea that workers are training the machines that will replace them is the most explicit version of a transformation affecting millions of jobs around the world, including sectors such as customer service, translation, programming, and data analysis. The difference is that in hand farms, the process is visual, immediate, and impossible to ignore.
For labor market experts, the phenomenon of factories in India is a harbinger of what could happen on a global scale. If the training of artificial intelligence by human workers is the bridge between current manual labor and the automated work of the future, millions of professionals across all sectors may be unknowingly contributing to their own replacement every time they use AI tools that learn from their actions. The difference between hand farms and corporate use of AI tools is merely the transparency of the process.
The ethical implications that the video of workers raises
The viral spread of the video raised ethical questions that the companies involved prefer not to discuss publicly. What is the acceptable limit for using humans as sources for AI training? Should workers be explicitly informed about how their data will be used? Are there regulations that should prevent companies from training systems that will replace their own employees? These questions still lack clear answers in virtually any country in the world.
The origin of the specific video has not been definitively determined, and no company has publicly claimed the practice. However, experts confirm that situations similar to those shown in the footage occur in factories in India and other countries, and the practice is expected to expand as the development of humanoid robots advances. The viral video may have made it impossible for the industry to continue operating without publicly discussing what it does, but social media outrage rarely translates into effective regulatory change. Workers continue to wear helmets, cameras continue to record, and artificial intelligence continues to learn.
A factory video in India shows workers training the AI that will replace them. Do you think this practice should be banned? Could the same happen to your profession in the future? Share your opinion in the comments.
Seja o primeiro a reagir!