The company Facebook has shared new details on the research that may ultimately lead to the creation of more advanced AI-technologies.
“Much of our work in robotics concentrates on self-supervision, in which systems learn directly from raw data (rather than from extensive structured training data specific to a particular task) so they can adapt to new tasks and new circumstances”, the research team wrote on their blog.
In particular, the team tried to teach a six-legged robot to walk without help.
“We have been able to design algorithms for AI and actually test them on a really challenging problem that we otherwise don’t know how to solve.”
Read more on the topic:
The hexapod begins its existence with no understanding of its surroundings. Due to the recursive function developed for self-improvement, the robot can gather information about its environment in order to optimize its behavior. The more experience the robot gains, the better it performs.
At first glance, nothing complicated, but we should consider the fact that the robot is expected to figure out not only its location and orientation in space but also to “feel” the impulses of the sensors located in the machine’s knees.
By optimizing the robot’s behavior and focusing on the task, the researchers from Facebook managed to teach the robot how to walk in just a few hours rather than days.
It would seem that the next thing to do for the hexapod is to go exploring but it’s not so easy to induce wanderlust in robots. Nevertheless, this is exactly what Facebook is going to do in the next stage of the research.
We want to remind you:
Subscribe to The Coin Shark news in Facebook: https://www.facebook.com/coinshark/