Term | Definition |
artificial intelligence (AI) | The computing field focused on developing computers that can perform functions normally reserved for humans: thinking, talking, seeing, feeling, walking, and learning from mistakes. |
cognitive science | The study and simulation of the human mind, or the area most commonly thought of when considering artificial intelligence. |
intelligent agent | A software robot (or bot), which is an intuitive assistant (agent) that can utilize knowledge based on past experience and predictions of likely future behavior. The agent can then use that knowledge to assist computer users. |
knowledge engineer | A programmer who specializes in building knowledge bases, each consisting of a set of rules that the computer will try to prove or disprove by asking a battery of questions. |
inference engine | The portion of an expert system that enables the system to draw deductions from the rules in the knowledge base in response to user input. |
genetic algorithm | A technology that applies a Darwinian (survival-of-the-fittest) method to problem solving, using a computer to create thousands or even millions of slightly varied designs, and then testing and selecting the best of them. |
mutation | A more random variation of the genetic algorithm programs which have evolved to another generation. |
crossover | A situation where the most successful mutations from genetic algorithm designs are mixed to produce a new generation of “offspring.” |
fuzzy logic system | An AI system that allows users to input “fuzzy” data, instead of demanding precise yes/no or numerical information. |
neural network | A system that simulates the physical workings of the human mind, and presents the ultimate attempt to model human intelligence. |
machine learning | A term that refers to an AI system that can learn from data. |
natural interface | An interface between humans and machines that uses speech recognition, natural language technologies, and other more human communication methods. |
natural language interface | An area of AI that pushes beyond speech recognition and web search, with the goal to have a machine that can read information on any topic, understand what it has read, and summarize what it has learned for the user. |
natural language processing (NLP) | The AI technology that enables a person to interact with a computer by either speaking or typing sentences rather than selecting commands. |
chatbot | A type of intelligent agent that uses AI. This type of computer program simulates intelligent conversation with the user via either spoken or typed inputs and responses (NLP). |
immersiveness | The quality of how real the simulated world feels, and how well it can make users accept it as their own and forget about reality. |
mental interface | A type of computer-human interface that uses sensors mounted around the skull to read the different types of waves or power spectrums the brain gives off, also called electroencephalogram (EEG) activity. |
robotics | The science of creating machines capable of independent movement and action. |
android | A simulated human. |
visual perception | Within robotics, the field working to enable a computer to understand what it sees with an electronic eye (camera). |
stereoscopic vision | Allows perception of depth and movement, simulated by two cameras. |
audio perception | Within robotics, the field working to enable a computer to hear sounds and understand what it is hearing, as in speech recognition. |
tactile perception | Within robotics, the field working to enable a computer to understand what it is touching. |
dexterity | Within robotics, the field working to enable a computer to have hand-eye coordination. |
locomotion | Within robotics, the field working to enable a computer to perform broad movements, such as walking. |
navigation | Within robotics, the science of moving a mobile robot through an environment. |