A better test than turing [News]

Abstract

The Turing test is a flawed metric for human-level AI. Can Winograd schemas do better? In June 2014, a computer program named Eugene was able to convince 33 percent of the humans it chatted with that it was a real 13-year-old Ukrainian boy named Eugene Goostman. The experience was part of an artificial intelligence contest held by the University of Reading, in England. Eugene was declared to be the first AI to have passed the Turing test𠄺 Turing's 1950 attempt to provide a framework for determining whether machines can think.