Passing the Turing Test

At the dawn of computing history, the brilliant scientist and World War II codebreaker Alan Turing began to uncover many of the fundamental principles of computer science. He thought deeply about computing and foresaw a time when computers would become much more powerful. In a famous 1950 paper titled “Computing Machinery and Intelligence,” Turing asked the question: Could machines ever “think?” If, indeed, machines could one day be able to “think,” how would we know?

In answer to this second question, Turing proposed what has become known as the “Turing Test.” This test involves a human interrogator, another human and a computer. An interrogator is placed in a separate room and exchanges messages with the computer and another human to determine which is which. Turing suggested that if the interrogator could not reliably identify which was the computer, “one will be able to speak of machines thinking without expecting to be contradicted.” Turing predicted that computers would pass the test by the year 2000.

While computers have triumphed against human opponents in several different areas, the year 2000 came and went without any passing the Turing Test. However, last month, precisely 60 years after the death of Alan Turing, a supercomputer officially passed the Turing Test at an event held by the University of Reading. The computer program, named “Eugene,” fooled 33 percent of the contest judges, barely exceeding the meager threshold of 30 percent required to pass the challenge. Eugene was programmed by a team from Russia to mimic a 13-year-old boy from the Ukraine. Some see this as a milestone for artificial intelligence, while others remain unimpressed.

Prior to this event, many clever “chatbot” programs have been written to simulate intelligent conversation. An early example of a “chatbot” program was ELIZA, a computer simulation of a psychotherapist written by MIT professor Joseph Weizenbaum in the 1960s. Although the program was based on simple pattern matching techniques, to the surprise of Weizenbaum, many users were emotionally drawn into the program. More recently, some companies have employed automated “dialogue systems” to provide online help. Some people who enter online conversations or who reply to certain emails have actually engaged a “chatbot” unawares. As computers become more proficient at passing the Turing Test, it will become more difficult to discern whether you are talking to a person or a machine.

Processing, not understanding
But what does the fact that a machine has passed the Turing Test really mean? Is this irrefutable evidence that machines are capable of having a mind, that they can “think”? The philosopher John Searle argued that even though computers can simulate human conversation, they will never have understanding. He argues his point by introducing the “Chinese room” thought experiment. The experiment imagines a person locked in a room who only understands only English. Messages written in Chinese are passed into the room and the person is given detailed instructions on how to give back certain Chinese symbols in response to the Chinese symbols he receives. Although the person inside the room does not understand any Chinese, it would appear on the outside that he does understand. Searle argued that a computer is essentially a symbol processing machine and it will never understand the way that humans do. In the words of a famous computer scientist, the late Edsger Dijkstra, the question as to whether machines can think is about as relevant as the question of whether “submarines can swim.”

The Turing Test is based on a faulty presupposition: that if machines can convincingly mimic human conversation, this makes them somehow like humans. Even though machines may now pass the Turing Test, the distinctive nature of humans created in the image of God will always set them apart from machines. The creation story clearly sets apart the creation of humankind, and we are certainly made up of more than the stuff of the earth. The vision of the dry bones in Ezekiel required more than flesh to bring them to life; they needed breath from the Lord. We are more than machines, and what we are cannot be captured by the Turing Test.

Even if computers are able to pass the Turing Test convincingly, we will lose something if we decide to replace human interactions with human-computer interactions. That being said, don’t be fooled: your next online conversation could actually be with a machine, perhaps one named Eugene.

  • Derek C. Schuurman is a Canadian currently living in Grand Rapids, Michigan where he is professor of computer science at Calvin University. Prior to arriving at Calvin he taught for many years at Redeemer University College and was a visiting professor at Dordt University. He currently holds the William Spoelhof Teacher-Scholar-in-Residence chair at Calvin. Besides his technical interests he is interested in faith and technology issues. He is the author of Shaping a Digital World: Faith, Culture and Computer Technology (IVP, 2013).

You just read something for free.

But it didn’t appear out of thin air. Writers, editors and designers at Christian Courier worked behind the scenes to bring hope-filled, faith-based journalism to you.

As an independent publication, we simply cannot produce award-winning, Christ-centred material without support from readers like you. And we are truly grateful for any amount you can give!

CC is a registered charity, which is good news for you! Every contribution ($10+) is tax-deductible.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *