Cognitive Science – The Artificial Intelligence Approach

A potential for emulating the human brain

Shift the focus of research to the realization of an intelligent agent

Aims to understand and build intelligent machines

Entities that act/ interact with an environment (i.

e. hands, legs, mouth)

Developed by Alan Turing, a concept that provides that basis on which all modern digital systems are designed: in its contemporary form, commonly referred to as a "finite-state" design or "sequential-system" design; could be employed to design and build an AI machine

Readily implemented within a machine with dedicated integrated circuit switches that can control the entire system

Digital computer organizations

Calculations, logical decisions, program sequence control; determines state transitions; makes cognitive decisions (cognitive manipulation)

Stores programs, results, temporary results, data; stores state definitions, external information, transition rules; facts, cognitive rules, cognitive methods

Sensor information, control of all external system elements; receives sensory information, provides control to external world changes; signals from external sensors and to external actuators, conversion to internal representation, conversion to action signals

Communication between other elements of the computer; communication with external world

Ultimate test of an IA, if a machine were able to pass the test in its most refined form, it would enable one to make the argument that the "intellectual" responses of that machine were indistinguishable from those of human being

Must surmount a significant number of hurdles

Information communication in this is bidirectional; upward flow produces coherent perception, but learned information in memory also flows downward

Completing a task, such as recognizing an object

Vast memory capacity

A functional relationship between the number of components in integrated circuits and industrial product cycles

Corresponds to 2^40 bits

Units of binary information

An intelligence that is beyond the human level

A collection of computational methods that have been modeled on the principles of biological evolution

Select actions on the basis of existing precepts - a "survival" mechanism (i.e.

IF car ahead is braking, THEN initiate braking)

Must keep track of precepts that they cannot follow continuously; thus, they maintain some internal state (i.e. agent negotiating a maze needs to keep track of map)

Actions depend on the goal to be achieved, in which the agent retains some goal information that describes desirable situations (i.e. when riding in a taxi, the goal might be a passenger's destination; agent must have some planning and searching algorithms to achieve this)

Enhanced goal-based agents, goals are not sufficient to achieve "high-quality" behavior, to these goals, we need to add a "quality factor"

An outgrowth of an idea from Turing, to build a learning machine and then teach it

A system may often include collections of different agents interacting with one another

The rational side of social science - where social is interpreted broadly to include human as well as nonhuman players, such as computers, animals, and plants (i.

e. the prisoner's dilemma)

The programmer must find a suitable representation of the information

Facts, rules; the programmer must specify a sufficient set of rules to define all the categories that the program must support

Algorithmic processes, the programmer must specify in advance the actions to be taken by the system for all combinations of inputs that may occur

Numerous examples come from biological systems where researchers seek to develop models for this behavior, bird migration, visual navigation, and predator-avoidance behavior in fish