Artificial intelligence researchers applied human learning techniques to a computer program and found that the program performed better than humans. According to them, the breakthrough will be helpful for humans in the future.

Earlier, researchers tested a computer program and found that it could understand and recognize approximately 50 handwritten languages, including Gujarati, Tibetan, Sanskrit and Glagolitic. According to the researchers, it was a successful test, which proved that machines are better than humans when it comes to recognizing the figures.

Three researchers who develop the new computer program said it is capable of recognizing handwritten drawings after viewing them for few times. The program also passed a Turing test, as per the researchers whose findings have been published in journal Science.

Joshua Tenenbaum, a professor from the Massachusetts Institute of Technology and co-author of the findings, said in a statement, “For the first time, we think we have a machine system that can learn a large class of visual concepts in ways that are hard to distinguish from human learners”.

The computer program, called the Bayesian program learning framework (BPL), proved that in some instances, it could even perform better than humans. In a test, humans had an average error rate of about 4.5% while BPL was with an error rate of just 3.3%.

The researchers said that during Turing test, the program performed well. In some cases humans could not find the difference between computer-drawn figures. They are now hopeful that the program may help to develop applications that rely on symbols.