Vancouver-based company Sanctuary AI is developing a humanoid robot called Phoenix that, when complete, will understand what we want, understand the way the world works, and have the skills to carry out our commands.
"The total long-term addressable market is the largest ever in business and technology history, which is the job market," says Jordi Rose, the company's CEO.
According to BBC News, dozens of other companies around the world are working on this technology.
In the UK, Dyson is investing in artificial intelligence and robots aimed at doing household chores.
Perhaps the best-known company in the market is Tesla, Elon Musk's electric car company.
It's working on the humanoid robot Optimus, which Musk says could be on sale in a few years.
"Ten years at the pace at which technology is moving now is an eternity," says Rose, who has experience in theoretical physics. "Every month, there are new developments in the world of AI that are like fundamental change."
Mainstream interest in AI exploded late last year when a powerful version of ChatGPT was announced. Its ability to generate all kinds of useful text and images has spawned competitors and a wave of investment in AI technology.
But developing an AI that allows a robot to complete useful tasks is a different and more difficult task.
Unlike ChatGPT and its competitors, bots have to navigate the physical world, and they need to understand how things in that world relate to each other.
Tasks that seem easy to many humans are major achievements for humanoid robots.
Sanctuary's Phoenix robot, for example, packed clothes into plastic bags in the back room of a Canadian department store in a pilot project.
"Handling the bags is actually quite challenging for the robots," Rose says.
One of the biggest challenges is giving the robot a sense of touch, so it knows how much pressure to apply to the body.
And there is still a great deal of work to be done to build a robot that can handle all the events that can happen in a busy home or workplace.
"You can't put a robot in an unstructured environment and then ask it to move without basically destroying things," says Professor Alireza Mohammadi, who founded the Robotic Motion Intelligence Laboratory at UCLA.
He points out that you can put an AI through millions of training scenarios, but there is always a chance that in the real world it will encounter something it has never seen before and react in an unexpected and potentially dangerous way.
Part of the problem, he says, is that humans have an intuitive understanding of context and consequences.
“Within ten years, we may have robots able to get around with some guidance, but not in completely unstructured environments,” says Professor Mohammadi. “But if these challenges can be overcome, could human-like robots begin to take over the jobs that humans do?” Humans now?