AI will learn new skills by browsing the video in fast mode

AI specialists from the US Army laboratory and developers of the University of Texas created the algorithm of artificial intelligence Deep TAMER, which will be able to learn new skills faster by browsing the video in an accelerated mode. A person who trains AI will observe the learning outcomes, praise or scold the robot, depending on the situation. The process of training itself resembles the training of dogs, and the speed with which AI will be trained in new “tricks” should at the same time grow at times.

Lead author of the research and technology developer Dr. Garrett Warnell explained that with the new AI training program, robots will be able to learn much faster, adapting to circumstances and adjusting to current tasks. In addition, he noted that this work will be an important step in the creation of fully autonomous devices that are able to independently make decisions and perform a number of complex tasks, not only in laboratories, but also in the field.

Now the robots learn for a long time, interacting with objects through trial and error. Only after a series of incorrect sequences of actions they begin to arrive more or less correctly, but during training the robot, for example, can completely break down, falling off a rock or drowning. A new method and people’s help will speed up training and help avoid fatal errors.