Skip to Main Content

Artificial Intelligence (AI) 2024 Guide: Glossary

AI resources in the library and links to U.S. government and international AI documents and organizations..

Glossary of Terms

Algorithm -- A mathematical equation or procedure used for solving a problem.

Artificial General Intelligence (AGI) -- A type of strong AI that is capable of matching human intelligence in a wide range of cognitive tasks and can teach itself to solve problems and complete tasks without human involvement.

Artificial Intelligence (AI) -- Technology that allows computers and machines to simulate human intelligence and problem-solving capabilities.

Artificial Super Intelligence (ASI) -- A type of strong AI that surpasses human intelligence and develops consciousness.

Bias -- Prejudice in favor of or against one thing, person, group, or way of looking at a situation compared with another, usually in a way considered to be unfair.

Computer Vision -- A field of artificial intelligence that uses machine learning and neural networks to teach systems get useful information from images, video, and other visual inputs.

Disinformation -- False information that is deliberately spread to mislead people.

Deepfake -- Fabricated hyper-realistic digital media, including video, image, and audio content. 

Deep Learning -- A subset of machine learning that uses multiple neural networks.

Expert Systems -- Computer systems trained on one specific topic and designed to make decisions on that topic in a way a human expert does.  

Fuzzy Logic -- A subfield of AI that allows systems to analyze vagueness and uncertainty or imprecise data using mathematical systems. These systems offer best possible solutions given the input. 

Generative AI -- An AI system that learns to generate more objects that look like the data used to train it.

Hallucination -- Incorrect or misleading results generated by AI. This occurs when the system has not been trained on the correct data, or when it has been trained on inaccurate or biased data. 

Large Language Model (LLM) -- A deep learning algorithm that can recognize, summarize, translate, predict, and generate text and other forms of content based on knowledge gained from massive datasets.

Machine Learning -- A branch of AI that focuses on the using data and algorithms to enable AI to imitate the way that humans learn, gradually improving its accuracy. 

Misinformation -- False or inaccurate information. Misinformation is often mistaken information.

Natural Language Processing -- A subfield of AI that enables systems to understand and interact with human language. These systems transcribe speech, analyze sentiment, summarize and translate, and power chatbots. Examples include Siri and Alexa, Google Translate, and speech-to-text apps. 

Neural Network -- A machine learning model that makes decisions similar to how the human brain does. It uses nodes to mimic the way biological neurons work together to identify patterns and solve problems.

Robotics -- The design, construction, and use of machines to perform tasks done traditionally by human beings.

Singularity -- A point in time where AI surpasses human intelligence and can no longer be controlled by humans.

Strong AI -- A theoretical form of artificial intelligence that equals or surpasses human capabilities.

Turing Test -- A 1950s-era test named for computer scientist Alan Turing that requires a human to interact with an other human and a computer in separate rooms and determine which participant is the human and which is the computer. 

Weak AI -- Current forms of AI that can automate tasks and analyze data.

Terminology adapted from:  American Psychological Association, BritanicaGoogleIBM, MIT, NVIDIAOxford LanguagesStanford, and TechTarget.