F. Hernández Quiroz and R. Morado (Mexico)
Turing machines, intelligence, problem solving, method, computational models.
The adoption of Turing machines as a model of intelligence can be motivated by some assumptions, both mathematical and philosophical. In a previous paper we talked about the assumptions related to the possibility, the necessity, and the limits of representing problem solving by mechanical means. In this paper we discuss general assumptions that have to do with considerations of method. Among the assumptions, we start with the idea that mechanical intelligence can be seen as the application of some method. Specifically, TM’s embody a method for information processing through syntactical transformation. Such transformation is achieved via following rules that specify a sequence of actions, and the specification of such rules must be unambiguous and have a certain indifference to time and the past. The specific assumptions about the method are that it is mechanical, finite, with identifiable results, disregarding in a certain sense time and space. We also discuss what might happen if the assumptions were to be rejected or weakened. We hope that tinkering with these assumptions will help to shed light on the import of alternative computational models.
Important Links:
Go Back
