Can Machines Think?
The concept of artificial intelligence was explored in the mid-1940s, but it wasn’t until 1950, when Alan Turing published his paper “Computing Machines and Intelligence” that we asked ourselves, “Can machines really think like humans?” This was the logical structure of Alan Turing’s 1950 paper “Computing Machines and Intelligence” in which he discussed how to build intelligent machines and how to test their intelligence. Because of these and other complications, Alan Turing proposed to bypass the question “Can a machine count” entirely.
In the end, Turing came to the conclusion that there is no better answer to the question of artificial intelligence yet, but certainly not ever. The challenge is an extension of the original question that Alan Turing was trying to answer, but would perhaps offer a high enough standard to define a machine that could “think” in the way we normally define a typical human. If a machine solves a computational problem that is almost impossible for a human to solve, the interrogator will know that the program is not a human and the machine will fail the Turing test. The Turing test itself is flawed because it doesn’t actually determine if machines can think, which is what it’s supposed to be.
The Turing test assesses whether a machine can fool another person into believing that it is talking to a person. The ELIZA example shows that Turing-tested machines can mimic human conversational behavior without thinking or thinking by following a simple (but extensive) list of mechanical rules. Mathematician Alan Turing believed that if a panel of judges could not distinguish between humans and machines, then machines would pass the test and be said to exhibit intelligent behavior or, in Turing’s view, thinking. If you want a machine to mimic the behavior of a human in some complex operation, you have to ask it how to do it, and translate the answer into a list of instructions.
While you can program machines with millions or even billions of different things, that doesn’t really make them think independently. If the question of whether machines can think is not addressed in terms of simulation, then it is possible to realize that answering the need to define and to go on to define the terms machine and mind is really important. Perhaps the most famous response to this argument came from John Searle in 2004, who developed a thought experiment he claimed to prove that machines cannot think.
