 This week, Boston Dynamics and other world's leaders in robotics unveiled their updated Atlas robot. And those images were startling for no other reason that they show how far robots have come over the course of the last 10 years. Here was a machine which more closely resembled the science fiction visions of Blade Runner and Terminator rather than overhyped toaster. And while the front line of automation in terms of work over the next decade or two will be in autonomous vehicles, algorithms and machine learning, this seems startling and surely it must have implications as to jobs, unemployment and where economies go from here. The reason why robots, at least so far, haven't lived up to some of those historic expectations is that they've been actually subject to a paradox. It's called Moravex Paradox. It's named after the technologist of the same name. The paradox is simple to understand. It says that machines counter-intuitively are very good at tasks that humans find difficult. They can be a grandmaster in chess. They can accurately model an earthquake or a nuclear weapon being deployed. But then they struggle at things we find easy or natural, like washing up the dishes or picking up a paperclip and putting it in an envelope. In short, it takes far more computational power to engage in sensory depth awareness, to understand and adapt to space and objects in real time. What Atlas shows is that finally Moravex Paradox may be coming to an end. The question has been whether machines could engage in fine motor coordination. Well, Atlas can do somersaults, so there's your answer. If progress over the course of the next decade is as rapid as over the last decade, whether robots can replace us in all manner of jobs is not a question of if, but rather when. Indicative of that was the DARPA Grand Challenge, which was in 2004 in the Mojave Desert. They said to the finest minds of their generation, if you can create a machine, a driverless car capable of navigating 150 miles in the Mojave Desert, you will win a million dollars. Huge sum of money. It turned out that nobody could do it. In fact, the best performing car couldn't even finish 5% of the course, barely completing more than 7 miles and many of the vehicles couldn't even move more than a couple of hundred meters. It seemed that driverless vehicles were decades away, that the idea itself seemed ridiculous that it should be put off for a generation or two. Does this technology underpin mass technological unemployment, declining living standards, increasing inequality between rich and poor? Or alternatively, is it the path to a different kind of society, where we all work less and have ever more leisure time, eventually as the technology improves more and more, moving to a world beyond work and scarcity altogether? Fundamentally, this is in tune with something that John Maynard Keynes said in 1930. He believed that within a century, we would solve the economic question of scarcity. That would then fundamentally mean we had to address the elementary question of what it is to be human, how to live wisely, agreeably and well. It turns out Keynes might well be right. By 2030, we may have to answer precisely that question. But if these technologies, like Atlas, are subordinated to capitalism and its social relations and the profit motive, that can never happen. Rather than paradise, humanity will be sent to hell.