Google DeepMind AI learns to creatively move around obstacles

From Engadget - July 10, 2017

The team wanted to see if simple rewards would work in a complex environment. They set up a virtual parkour course with drops, hurdles and ledges and set a reward for forward progress. At its most basic level, the system was as follows: the faster the AI moved across the terrain, the greater the rewards. Additional incentives and penalties were added for more complex programs.

You can see the full results in this video; all of the stick figure's navigation was taught via reinforcement learning. The AI used a trial and error system to figure out how to move forward as fast as possible without "terminating."


Continue reading at Engadget »