Learning to do robotics has easily become exciting. But doing this job has always required a lot of knowledge. What if there was an easier way for normal people to teach these systems to do things that humans can do?

At the moment, two of the most popular ways to learn are through copying and reinforcement. First, you have to control the robot and teach it how to do work. Second, you have to train a system on millions of pictures. Several people who study robots are looking into a method that trains a robot by making it watch a person do a job. A team at Carnegie Mellon University showed the Wild Human Imitating Robot Learning, or WHIRL, an algorithm that teaches a robot by making it watch videos. In a news report, Shikhar Bahl, a student at a robotics school, said that copying is a good type of learning.

So, it’s easy to say that these new inventions will be very helpful in the times ahead.