Skip to main content

Artificial Intelligence Still Lags Behind Humans, Fails Turing Test

In spite of great advancements in artificial intelligence and robotics, computers are still not good enough to pass themselves off as humans in a text based conversation, according to a study done by team of scientists at University of Reading.

Though the machine, named Elbot, came closer with a success rate of 25 percent in persuading human beings that they were conversing with other humans, it couldn’t pass the standard of 30 percent, as set by the famous British mathematician Alan Turing.

The scientists conducted an experiment on Sunday, codenamed ‘Turing Test’, during which six Artificial Conversational Entities (ACEs) attempted to fool human interrogators.

According to a spokesman from the University of Reading, “During the tests, all of the aces managed to fool at least one of their human interrogators”, but none could achieve the set benchmark.

Professor Kevin Warwick from the University’s School of Systems Engineering said, “This demonstrates how close machines are getting to reaching the milestone of communicating with us in a way in which we are comfortable.”

Despite failure, the programmers of Elbot were awarded GBP1,760 prize for their achievement.

Désiré Athow

Désiré has been musing and writing about technology during a career spanning four decades. He dabbled in website building and web hosting when DHTML and frames were en vogue and started writing about the impact of technology on society just before the start of the Y2K hysteria at the turn of the last millennium. Following an eight-year stint at where he discovered the joys of global tech-fests, Désiré now heads up TechRadar Pro. Previously he was a freelance technology journalist at Incisive Media, Breakthrough Publishing and Vnunet, and Business Magazine. He also launched and hosted the first Tech Radio Show on Radio Plus.