Just as we've become accustomed to algorithms constructed by humans but unleashed on the world where they can morph and evolve, robots have started to learn and function on their own once released by their human overlords. "Robots that have the ability to 'learn' and do specific tasks are nothing new," TechCrunch reminds us, "but truly autonomous models are still a thing of the future."
If efforts from the Tokyo Institute of Technology are successful, that future could be rapidly approaching. There, engineers and scientists are "working on a robot that's supposed to be able to learn, adapt to new situations and act in a human-like way someday," according to TechCrunch.
The video embedded below shows HIRO, a robot that uses SOINN (Self-Organizing Incremental Neural Network) to learn from new situations. "This robot remembers only basic knowledge, and it can apply that knowledge to its immediate situation," one of the programmers responsible for HIRO explains in the video. "If it doesn't know enough, it stops, and reacts by saying, 'I can't do this because I don't know how.'" But what makes this robot especially interesting (or frightening, depending on one's perspective), is that, once it recognizes it doesn't know how to do something, it can tap into the Web or the neural network of other robots like it. From there, it will learn, retaining the information it might need to complete similar tasks in the future.
Watch other Videos on the Technology Channel.