There are always jokes about how robots could one day replace humans at their jobs. However there is a simple solution to this, and that is to simply not make these robots in the first place. Unfortunately it looks like it’s not as straightforward as that, thanks to the efforts of NVIDIA and a team of researchers who have developed a deep learning system that basically lets robots learn how to complete tasks simply by observing a human.
The idea isn’t so much about robots taking over our jobs, but more about letting humans and robots communicate with each other better and allow both of them to work side-by-side seamlessly at the workplace. According to the researchers, “For robots to perform useful tasks in real-world settings, it must be easy to communicate the task to the robot; this includes both the desired result and any hints as to the best means to achieve that result.”
They add, “With demonstrations, a user can communicate a task to the robot and provide clues as to how to best perform the task.” This means that robots can be used to observe how a human does a task to complete it, after which the human operator can leave the robot to its own devices while they go off and do something else. You can check out the efforts of the research at NVIDIA’s website, or watch the video above to see it in action.