Trends4Tomorrow (March 3, 2009) — In his landmark 1950 paper, “Computing, Machinery & Intelligence,” British mathematician Alan Turing theorized, “If, during text-based conversation, a machine is indistinguishable from a human, then it could be said to be ‘thinking’ and, therefore, could be attributed with intelligence.”

He further suggested that if a machine took part in a series of five-minute-long conversations with people and was able to convince more than 30 percent of those people that it was a human being, it could be deemed to have passed the so-called “Turing Test.”

Recently, at a major artificial intelligence competition at England’s University of Reading, machines for the first time came close to imitating human communication and successfully passing the Turing Test. 

In fact, a program called Elbot, created by programmer Fred Roberts, successfully fooled 25 percent of the human interrogators, coming extremely close to the 30 percent Turing Test threshold set in 1950. 

Meanwhile, five of the competing programs, including Elbot, were assessed by those who correctly identified them as computers, to have achieved 80 to 90 percent of the communication ability of actual humans.

Over the next few years, we can expect this functionality to start showing up in commercial applications such as educational “bots,” office automation systems, and e-commerce.