A human cleaner has a really big advantage over a vacuum: it can move things out of the way. That doesn't seem like a huge problem, but the vast majority of robot vacuums lack arms. It can't just move a chair or get a hoover nozzle into a tight spot to clean up dust. There's also the unknowns to think about that it won't have mapped on its route, like toys, charging cables left trailing across the floor or even pet messes that can be spread around the house. That's not to say that robot vacuums can't overcome some of the issues.

Noland Arbaugh's life changed in a fraction of a second in June 2016. He was a 22-year-old student, working at a kids' summer camp in upstate New York, when he went swimming in a lake. He can't tell me exactly what happened, but thinks one of his friends must have accidentally struck him very hard in the side of his head as they ran into the water and plunged beneath the surface. When he woke up face down in the water, unable to move or breathe, Noland immediately knew he was paralysed. He felt no fear at all, he says. "You never know what you're going to do in those high-stress situations. I found out that day that it's hard to shake me. I am very, very calm under pressure." Elon Musk would ultimately turn this quality to his advantage when, after nearly eight years of being quadriplegic, Noland agreed to allow the world's richest man to implant an electronic chip into his brain. In January 2024, Noland became the first human recipient of a brain-computer interface (BCI) developed by Musk's company, Neuralink. If it worked, it would allow him to control a computer using only the power of his mind. Only four months after he first heard about Neuralink, Noland was on an operating table, with a purpose-built robot poised to insert the N1 chip into his motor cortex. The stakes could not have been higher for him: he was risking infection, haemorrhage and brain damage. "My brain is the last part of myself that I really feel I have control over," he tells me from his wheelchair at his kitchen table in Yuma, Arizona.

A robot that can quickly dig a bottle of milk out of the depths of a cluttered fridge has been created by a team of British scientists. The artificial intelligence does this by combining automated planning -- based on images taken of the objects -- with trial-by-error learning. Alongside rearranging fridge contents, the same setup could be applied to allow robots to perform various similar tasks -- such as locating items in a warehouse. When reaching into a crowded fridge, it may sometimes not be possible to remove, say, a desired pint of milk without first shuffling other items in the way around first. While this kind of operation may come easily to us, such a complex task for a robot -- requiring it to devise a chain of separate movements leading to the end goal which can be time consuming to calculate.