Portuguese 
English 
Spanish 
5 min of reading 
0 comments 
In the U.S., people are paid to do household chores with body-mounted cameras, generating data that trains robots and creates a new type of work in the age of AI.
In Los Angeles, an apparently mundane scene — someone washing dishes or cleaning the kitchen — is being transformed into something much larger: a database for training robots. Hundreds of people are being hired to perform household tasks while wearing body-mounted cameras, recording every movement from a first-person perspective, according to a report by the Los Angeles Times. What seems like a common job has become part of a new emerging economy, where everyday activities are converted into valuable information for the development of physical artificial intelligence. This phenomenon marks an important transition: from AI that understands text and images to AI that learns to act in the real world.
A new type of work: the “digital gig” of robotics
The model follows the logic of the so-called gig economy, but with a crucial difference. Instead of driving or delivering food, these workers perform household tasks while generating data.
Technology companies pay amounts that can reach around $80 for a few hours of recording. In exchange, they receive something extremely valuable: detailed records of how humans interact with objects, make decisions, and perform everyday tasks.
-
A “silent skill” is allowing Brazilians to earn up to R$ 22,000 per month without a degree and become indispensable for companies that rely on millions of data to survive.
-
Researchers at the Toyota Research Institute found that if a human uses robotic arms to flip a pancake 300 times in an afternoon, the robot learns to do it on its own the next morning, and this is currently the most promising method to solve the biggest bottleneck in modern robotics.
-
Goodbye iron: a common item in households is starting to lose space to technology that smooths clothes in minutes without an ironing board and with less energy consumption.
-
Antarctica reveals an unusual clue high in the Hudson Mountains, and what appeared to be just an isolated rock began to expose a secret hidden under the ice for ages.
This type of activity is being described as a new type of occupation — a blend of manual labor with data production — that could grow rapidly as robotics advances.
Why robots need to observe humans to learn
Unlike language models, which can be trained with texts available on the internet, robots face a different challenge: understanding the physical world.
For a robot, simple tasks like washing a dish involve multiple variables:
- how to hold the object
- how much force to apply
- how to react to wet surfaces
- how to adjust movements in real-time
These nuances are difficult to program manually. Therefore, companies are adopting learning by demonstration, where machines learn by observing humans.
Body-mounted cameras: how data is captured
Workers use devices that capture action from a human point of view. This includes cameras mounted on the head, chest, or even hands.
The goal is to record not only what is being done, but how it is being done. The first-person perspective allows algorithms to analyze:
- movement trajectories
- coordination between hands
- interaction with objects
- execution time
This type of data is essential for training robotic systems that need to operate in real and unpredictable environments.
From the kitchen to the lab: the path of data
After recording, the data is processed by artificial intelligence systems that identify behavior patterns. These patterns are then used to train robots or simulations.
The process involves:
- motion extraction
- action modeling
- reproduction in a virtual environment
- transfer to physical robots
This cycle allows machines to learn complex tasks without the need for detailed programming for each situation.
The global race for humanoid robots
The growth of this type of work is directly linked to the global race to develop robots capable of operating in human environments.
Companies like Tesla, Google, and various startups are investing billions to create robots that can perform household, industrial, and assistance tasks.
The ultimate goal is to develop machines that can operate in homes, hospitals, and factories with autonomy and safety. In this context, the data generated by humans becomes a strategic resource.
Homes become laboratories for artificial intelligence
One of the most striking aspects of this trend is the transformation of domestic environments into data collection spaces.
Kitchens, living rooms, and service areas begin to function as real scenarios where AI learns. Unlike controlled laboratory environments, these spaces offer variability and unpredictability — essential factors for robust robot training. This enhances the quality of the data and brings learning closer to real-world usage conditions.
The economic impact: physical data as a new commodity
The collection of data has always been one of the pillars of artificial intelligence. However, until recently, most of this data was digital. Now, a new category emerges: physical data. Movements, interactions, and human behaviors are gaining direct economic value.
This phenomenon could create a new market, where people are compensated not only for the work itself but for the informational value of their actions. In the long term, the development of robots capable of performing household tasks could profoundly alter the dynamics of work.
If machines can learn accurately how humans perform these activities, repetitive functions could be automated. At the same time, new types of jobs may arise, such as operators, trainers, and supervisors of robotic systems. This movement does not necessarily eliminate human work, but transforms it.
Between opportunity and questioning
Although the model represents a new source of income, it also raises important questions. The use of cameras in the home environment involves concerns about privacy, data security, and ethical boundaries.
Moreover, the amount paid for the work is still relatively low compared to the economic potential of the systems being trained. These points should gain relevance as the practice expands.
What is happening in Los Angeles may seem like an isolated experiment, but it represents something larger. It is a change in how artificial intelligence learns and evolves. By transforming simple tasks into structured data, technology brings machines closer to the human capacity to interact with the physical world.
The beginning of a new phase of artificial intelligence
The evolution of AI has always depended on data. Now, with the expansion into the physical world, this dependence intensifies.
The work of ordinary people, performing everyday tasks, is helping to build the next generation of robots. And this could redefine not only technology but also the very concept of work. What is today a “gig” may become, in the future, a central piece of a new economy based on interaction between humans and machines.
Seja o primeiro a reagir!