Can Machines Think?
The purpose of this paper is to summarize and contrast two opposing answers to the question "Can machines think?" or variations of the question, and to offer a way of thinking about intentionality that lends itself to a more straightforward approach to the problem. I find that when the dust settles after the standard arguments for and against the prospects of "strong" Artificial Intelligence, the problem that remains unsettled is the issue of intentionality. Searle's thought experiment in the Chinese room, together with the reactions it incites from his critics, provide a good framework for seeing why intentionality is the key. First, I will try to give an account of the Chinese room argument and various arguments against it, then I will be try to show why intentionality is so important. Finally, I will offer my personal opinions on intentionality and the prospects for creating artificial minds.
In his 1950 essay, Alan Turing proposed to answer the question "Can machines think?" by instead considering whether or not that machine can do well in an imitation game. This game involves an interrogator, the machine in question, and judges. Roughly, the machine must answer the questions posed to it by the interrogator sufficiently well as to fool the judges such that they cannot correctly determine whether they are witnessing communication between human and human or between human and machine. If the machine is successful, it passes the Turing Test, and we should consider it to be a thinking thing. For Turing takes the stance that mental processes could in theory be replicated, provided that the relevant programs can be run to sufficiently mimic the computational process.
This provides the foundation from which much of the debate progresses. In the Chinese room argument, Searle demonstrates the limitation of such a test, by finding an example where the observers are convinced that there is understanding going on, and yet it is clear that there is no...
View Full Essay