Startup Shift offers free house cleaning in exchange for video data for robots

Startup Shift has launched an unusual project in New York: it offers city residents completely free cleaning. In return, the company asks for one thing: to film the cleaners' entire working process from start to finish.

Free cleaning, paid for with video data

The project sounds appealing: skip the hassle, let professionals tidy up your home, and not pay a single ruble out of your pocket. In exchange, Shift gets hours of video that capture every nuance of the work: how dishes are scrubbed, how tables and floors are wiped, how different surfaces are handled. The company already plans to expand to London, San Francisco and other cities, gathering data volumes in each region.

Robots urgently need training data

Behind this noble idea lies a pragmatic calculation. Tesla Optimus, Boston Dynamics, Figure AI and dozens of other companies building household robots are hitting the same wall: a shortage of high-quality training video. Teaching a machine to wash dishes takes not hundreds but hundreds of thousands of hours of real-world footage. That is exactly what Shift is collecting:

• Real homes with unpredictable layouts and difficult access
• Different types of dishes, tables and floors, not laboratory standards
• Natural human movements that vision algorithms cannot model
• A huge volume of data shot in varied climate and household conditions

Such video is expensive if you try to buy it on the open market. Shift found a way to get it almost for free.

Why this isn't so simple

It sounds easy, but the devil is in the details. First, video is only half the battle. A robot must learn to interpret the footage: to understand what exactly the cleaner is doing and why. That requires a colossal amount of annotated data and powerful neural networks.

Second, privacy questions arise. Is it acceptable to film inside a person's home? What if relatives end up in the frame? Windows facing neighbors? Personal belongings? Shift has stayed silent on the subject so far, but sooner or later these questions will become critical, especially in Europe with its GDPR.

What this means

We are on the threshold of a real revolution in household robotics. Companies are no longer waiting for ideal laboratory conditions; they are going into the real world and collecting data there. Shift is simply a pioneer of this trend. If the model works, many companies will rush to copy it. And then ordinary people may get free household services in exchange for their work becoming a teacher for future robots.