A recent morning I played the main door of an elegant two -story house on Redwood City, California. In seconds The door opened a robot without facedressed with a beige Mameluco well adjusted to his waist and his long legs.
This slender humanoid welcomed me talking to what seemed like a Scandinavian accent and I proposed to strengthen our hand. When we touch our palms, he told me: “My squeems are strong.”
When the homeowner, a Norwegian engineer named Bernt Børnich, asked for some bottled water, The robot turned, walking in the kitchen and opened the refrigerator with one hand.
Artificial intelligence already handles cars, writes essays and writes computer code. Now the humanoids, Machines built to look like human beings And driven through AI, they are about to settle in our homes to help us with everyday tasks.
Børnich is executive director and founder of the emerging company 1X. Before the end of the year, this startup hopes to place your robot, NEO, in more than 100 homes of Silicon Valley and other places.
The Børnich company is between dozens of companies that plan to sell humanoids to homes and businesses. According to the investigation firm PitchBookwhich analyzes the technological industry, since 2015 investors have been allocating 7,200 million dollars to just over 50 startups.
Last year the fury for humanoids reached a new peak when investments exceeded 1.6 billion dollars.
This without counting the other billions that Elon Musk and Tesla, their electric cars, are investing in Optimus, A humanoid that began to build a long time agoexactly in 2021.
Entrepreneurs such as Børnich and Musk believe that one day humanoids will do much of the physical work that people now do, including domestic tasks such as cleaning countertops and dishwashing dishwasherdeposit works such as classifying packages and manufacturing work such as building cars in a assembly line.
The simplest robots – small robotic boots and autonomous wheelbarrows, for example – share with people for a long time the workload in deposits and factories.

Now companies are committed to the machines can make a widest range of tasks imitating the way in which human beings walk, bend, turn, reach, grab things and carry out their tasks in general.
Since houses, offices and deposits have been built for human beings, these companies argue that Humanoids are better equipped to develop in the world that any other robot that had a non -human form.
For years they have circulated online videos that show Notable skills Of these machines, but often remotely guided by human beings. And simple tasks such as loading a dishwasher are anything but simple for them.
Gestation in Silicon Valley
I first went to the 1x offices in Silicon Valley almost a year ago. When he entered the office A humanoid called Eve opening and closing the doorI could not avoid the feeling that this eye robot as dishes was actually a disguised person.
Eve moved on wheels, not legs. Even though it still seems human. I thought about The sleepingWoody Allen’s comedy of 1973 full of robotic butlers, and in the robotine of The supersonic.

The company’s engineers had already built neo, but The robot had not yet learned to walk. A preliminary version of the company was hung on one of the walls of the company.
In 2022, Børnich connected to a zoom video call with an ia -called Eric Jang researcher. They didn’t know each other until that moment.
Jang, who is now 30 years old, worked in A robotics laboratory at the Google headquarters In Silicon Valley, and Børnich, who is now 42, years, directed a startup in Norway called Halodi Robotics.
A potential investor had asked Jang to collect information about Halodi to see if it was worth investing there. Børnich looked exhibiting Eve.
It was something he had dreamed from building since his adolescence, inspired, like Many robotics expertsfor science fiction (your favorite movie: Blade Runner).
“I saw a hardware level that I did not think it was possible,” says Jang.
The potential inverter did not invest in Halodi. But soon Jang convinced Børnich to join forces.
Jang was part of a Google team that taught new skills to robots through mathematical systems called Neural networksthrough which they learn based on data that represent real world tasks. After seeing Eve, Jang suggested to Børnich to apply the same technique to humanoids.
The result was a transatlantic company that they renamed as 1X. The startup, which has grown up to about 200 employees, now has more than 125 million dollars in financing investors such as Tiger Global and OpenAi.
Approximately six months after meeting Eve, when I returned to the company’s laboratory he received a neo that was walking.
They had taught him to move fully in the digital world. Simulating real world physics in an environment similar to a video game, they had been able train a digital version of the robot to stand up, in balance and that, over time, take steps.
After months of training that digital robot, they transferred everything they learned by him to a physical humanoid. If I brought my way, Neo stopped and surrounded me to continue moving. If he pushed him in his chest, he kept standing. Sometimes it struggled or I didn’t know what to do. But I could walk through a room like people.
“All this is learned behavior“, Explains Jang, while each step of Neo hits the ground.” If we put it in any environment, you should know how to do this. “
Domestic tasks
When I returned to Børnich’s house a month later, Neo was starting to have problems with the stainless steel door of the refrigerator. His Wifi connection had fallen.
But once the hidden technician restarted it, he was able to guide the robot without problems in his small task. Neo reached me a bottle of water.
I also saw him load the washing machine beckling carefully to get the clothes from the basket. And while Børnich and I chatted out of the kitchen, The robot began to clean the countertops. Everything was done by remote control.
Even commanded by human beings, Neo could fall a cup or could have difficulty finding the right angle when trying to throw an empty bottle to a garbage tacho under the pool.
While they have improved a lot in the last decade, Humanoids are not yet as agile as people. Neo, for example, cannot lift your arms above your head.
While guiding Neo in domestic tasks, Børnich and his team can collect data -through cameras and others Sensors installed in Robot himself– They show how those same tasks are done.
1x engineers can use such data to expand and improve the skills that Neo has so far.
Just as Chatgpt can learn to write academic works by analyzing selected texts on the Internet, A robot can learn to clean windows identifying certain guidelines along hours of digital video.
Most humanoid projects, including the Optimus of musk and similar initiatives such as Apptronik y Figure AI, are oriented to Design humanoids for deposits and factoriesbased on those strictly controlled environments will be simpler for the navigation of robots.
For now, 1X It has not yet set a price for these machines that produces in its own industrial plant in Norway.
Goodbye to the human employee?
Manufacturing a humanoid like Neo costs approximately the same as manufacturing a small car: tens of thousands of dollars. “We want each home to provide us with their data under their own terms,” said the engineer.
Using that data, Børnich hopes to create A humanoid capable of performing virtually any domestic task. This implies that Neo could replace those who earn life by cleaning houses.

But that is still years away, at best. And because of the growing shortage of personnel who are responsible for cleaning housing or care of older people and minors, organizations that represent this workforce they see with good eyes The rise of new technologies that work effectively at home, provided that companies such as 1x produce robots that work well with human workers.
“These tools could facilitate some of the most expanding, demanding and dangerous works, and allow human resources to concentrate on things that people can only solve”, Comments Ai-Jen Poo, president of the National Alliance of Domestic Workers, which represents house cleaning staff, home care and baby sitters throughout the country.
Other humanoids that I have known can be intimidating. Neo, less than 1.65 meters high and 30 kilos, is not. However, me I wondered if I could hurt a pet, or a creaturewith a fall.
Will people let this machine enter their homes? How quickly will your skills improve? Can free people from their daily chores? These questions have no answer yet. But Børnich continues.
Translation: Román García Azcárate.