Tech companies are more than willing to film you doing homework

This week, an introductory AI tutorial called Shift said it would clean New Yorkers’ homes for free. It has plans to expand to other cities, including London, and looking around my house, I find it interesting.

But there is a catch. There is always a catch.

In exchange for the cleaning, Shift wants pictures of its cleaners at work: washing dishes, towels, napkins, tables, mopping floors. It wants everything. A video of the boring household chores we could happily use if we could – is that the robotics industry is rushing to train machines to do them in order to sell us something to do them for us.

It’s harder than it sounds. Unlike chatbots, graphic designers, and other AI tools that have exploded in recent years, robots have to deal with the world. This means understanding space, movement, energy, friction, strange shapes and materials, subtle lighting, and everything else that humans – and other species – tend to grasp instinctively. This is why things that are easy for us, like folding clothes, picking an apple, or pouring water, have proven to be very difficult for robotics experts to adapt.

Training machines to do those things is based on data. More about that. Text, images, and videos can be easily removed from the Internet for companies. And most of the time they didn’t give back to the people who made them. The animal world is hard to beat, and it’s hard to beat quietly without paying. This means that access to high-quality data is a major challenge for companies developing physical AI. It’s a lucrative opportunity, so companies like Shift are starting to develop.

They are not alone. In India, recent reports revealed that domestic service platform Pronto has been using customers’ homes as a source of AI training for household tasks such as cooking, cleaning, and laundry. Pronto says it only takes photos when customers walk in — it’s not clear what customers get, other than a copy of the video — but the practice still caused confusion in the market, with rival startups. to be stuck they have never written in-house to train AI and have no intention of doing so.

Some initiatives focus on data collection efforts. Created by Silicon Valley Human Archivefor example, they hope to partner with companies like Pronto and have gig workers record their events using unadorned camera cups. These helmets collect images from the wearer, exactly the type of “egocentric” or robotics companies that have to teach machines how people navigate places. Meanwhile, Shift also works directly with consumers, and says it has paid thousands of people in 15 countries to log their activities through its app.

Some companies are skipping useful tasks together. Instead, workers are paid to complete the same tasks over and over again while cameras and sensors record every move. Therefore data fields they were designed to make physical activities – napkins, collecting cups, carrying boxes – become AI training objects important enough to pay people to create them.

And some data is generated by robots already in the world. Despite the hype, real machines are still a long way off – hence the need for all this data – but companies are eager send the goods anyway. They will use data from customers’ homes to optimize sales. Many companies rely on remote workers to step in when the robots are shut down. They will use that data again.

Of course, the practice of trading data for a valuable commodity is not new. Companies have been offering discounts, conveniences, and free services to access your data for years, from loyalty cards and cookies to dashcams, insurance apps that track people’s driving, and more. Very bad TV that only shows commercials.

What’s new is the type of data companies are willing to pay for. For now, that means maybe letting someone in a shiny hat clean your house for free so that, eventually, a company can sell you a robot to do it for you.