Alibaba and Microsoft created artificial intelligence that exceeds a person in a test to understand what has been read

Alibaba and Microsoft created artificial intelligence, which went around a person in the Stanford University test to understand what has been read. This is written by Bloomberg.

This test is one of the most authoritative ways to evaluate artificial intelligence, the newspaper writes. A test of more than 100 thousand questions is based on approximately 500 articles “Wikipedia”. The answers to the questions could be explicitly indicated in the articles on the portal, and flow out of the read fragment of the text.

The result of a person in this test is 82.304, the system created by Alibaba gained 82.440, and Microsoft - 82.650.

As Lo Si, a senior language researcher in Alibaba Institute, these results suggest that machines can answer objective questions with great accuracy, for example, to the question of rain.

“In the future, artificial intelligence in the future will be able to use in the field of service, when creating museum teaching aids, online interaction with patients on medical issues,” Luo See said.