The Ultimate AI Glossary
Learn all about the various terms that are used in the AI industry.

The ultimate AI glossary


Algorithm It is a set of rules or commands provided to a computer so that it can complete a specific task.

AlphaGo It is a computer program that plays the board game Go. It is recognized as the first computer program to beat a professional human player at the board game Go.

Analogical Reasoning The ability to solve problems with the use of analogies and by comparison of past experience.

Artificial General Intelligence (AGI) It is also known as strong AI. AGI is a type of artificial intelligence that is considered to be human-like. It is currently in its preliminary stages of development.

Artificial Intelligence It is a subset of computer science where computer systems are programmed to perform tasks with the similar intelligence of a human. This includes decision-making, classification of objects, recognition of speech and its translation, etc.

Artificial Narrow Intelligence (ANI) This is also known as weak AI. This type of intelligence can only focus on a single task at a given time. This is the current form of AI that we use or deploy in our day-to-day activities and processes.

Artificial Neural Network (ANN) This is a network that is modeled to imitate the human brain. It is done so by creating an artificial neural system using pattern-recognizing algorithms. These algorithms learn, interpret, and categorize sensory information.

Artificial swarm intelligence It is a computer program that coordinates many individual technological systems to work together as a group. It mimics the behavioral structure of animals that work in swarms to arrive at an optimal solution.

Autonomous This is the ability of a machine or system to behave independently without intervention from a human in order to function well.


Backpropagation It is short for backward propagation of errors. It is a method in which neural networks are trained. In this method, the system’s first output is compared to the desired output and then it is adjusted until the difference in between the actual output and the desired output is reduced or becomes minimal.

Backward chaining This is a method where machines work backward from the desired output to find any data that validates the desired goal.

Bayesian networks It is alternatively known by various names such as Bayes network, Bayes model, belief network, and decision network. It is a probabilistic graphical model that represents a set of variables and their dependencies. It is used to build models from data and/or expert opinion.

Big data It refers to large amounts of structured and unstructured data that is too complex to be handled by standard data-processing software.

Black box It is a complex neural network or program used for used to check whether the output of a program is as expected, in accordance with the inputs. The term “black box” is used because the actual program with its algorithms and decision-making processes being executed are not examined.

Bot It is an autonomous program that has the capability to interact with other computer systems, programs, or human users. These are usually supervised directly or indirectly by humans.


Case-based Reasoning (CBR) This is an approach to problem solving using knowledge from solving problems in the past and applying that solution to the current problem.

Chatbots A chat bot is an extension of bots. It can converse with a human user using voice or text commands. This is predominantly used in industries where customers look for ease of communication.

Classification This is an algorithm technique that enables machines to bucket similar data points together and assign relevant categories to them.

Clustering This is an algorithm technique that enables machines to group similar data into larger data categories.

Cognitive computing It refers to a computerized model that mimics human thought processes using data mining, NLP, and/or pattern recognition. This technology is based partly on artificial intelligence and signal processing.

Computational learning theory This is a sub-field of artificial intelligence. This is a study on how to design computer programs that are capable of thinking and is aimed to identify the computational limits of machine learning.

Computer program It refers to a collection of instructions that is provided to a computer to perform certain tasks.

Computer science It is the scientific study of the use of computers and its principles.

Computer vision A term used when a machine processes visual input from image files (JPEGs) or camera feeds.

Convolutional Neural Network (CNN) It is a type of neural network or deep learning algorithm that created to analyze, classify, and cluster visual imagery by using multilayer perceptrons and be able to differentiate one from the other.

Consciousness It is a state of subjective experience or awareness.

Cyborg Less popularly known as ‘Cybernetic Organism’. It refers to a hybrid of human and machine. Think Terminator.


Data It is a digital collection of information of all kinds.

Data mining It is the process of identifying patterns in large sets of data in order to establish problem-solving relationships. It involves various methods that intersect at machine learning, statistics, and database systems.

Datasets It is a collection of data sets that are related. It is composed of various separate elements. These elements can, however, be manipulated as a unit by the computer.

Data science Interdisciplinary scientific field on processes and systems to extract knowledge or insights from data in its various forms.

Deep learning It is a machine learning technique that teaches computers to mimic the way a human mind learns using classification techniques.

Deep Neural Network (DNN) It’s a type of machine learning model. It is an artificial neural network that has more than two layers in between the input and output layers.

Dimensional reduction Process of reducing the number of random variables by obtaining a set of principal variables via feature selection and/or feature extraction. It is popular in applications such as image processing, video segmentation, and speech recognition.


Embodied AI It refers to the idea that intelligence is as much a part of the body as it is a part of the brain. It involves equipping software with a physical body and exploring how that body fits into real world environments.

Explainable AI It’s a type of AI that is able to explain its decision-making process.


Few-Shot Learning It refers to the practice of feeding very small amount of training data to a learning model in contrast to the normal practice of using a large amount of data.

Forward chaining It is the logical process of reducing complex tasks to multiple simpler tasks that are performed in a sequence. It is used to infer or draw insights from known data by moving forward using defined conditions and rules.

Friendly Artificial Intelligence (FIA) It refers to the artificial general intelligence whose values are aligned with that of our own.


General Intelligence It is the capability of an entity to achieve any goal. This is different from artificial general intelligence which refers to the ability of the entity to accomplish a cognitive task at the same level as humans.

Generative adversarial networks (GAN) It is a type of neural network that can take elements of photographic data to create realistic-looking images of people, places, objects, or animals.


Heuristic It is a solution-based problem-solving technique in computer science that is designed for optimal and quick outcomes.


Image recognition The ability or process of an artificial entity to recognize or detect an object or a feature of an object in an image or video.

Inductive reasoning It is a method in which the artificial entity uses data as evidence to create rules.

Intelligence It is the ability of an entity, both human and artificial, to accomplish complex tasks.


Long short-term memory networks (LSTMs) It refers to recurrent neural networks that are capable of learning long-term dependencies. It can both process single data points and entire sequences of data.


Machine learning (ML) This is a subset of AI focused on developing programs that perform specific tasks without too much human interference and mostly relying on patterns and inference.

Machine translation It is an application of natural language processing. It is used for language translation in text-based and speech-based conversations.


Narrow Intelligence It is the ability of an artificial entity to accomplish very specific tasks. The current applications of AI falls within the realm of narrow intelligence.

Natural language processing (NLP) It a type of machine learning algorithms that help process, interpret, and analyze human language using natural language data.

Neural networks It refers to computer system that are modeled after the human brain.


Pattern recognition It is the ability of a computer system to automatically recognize patterns in data.

Pruning It is a search algorithm used to remove undesirable or incorrect solutions to a problem from an AI system. This aids in reducing the number of decisions that the AI system must make.


Recurrent Neural Network (RNN) It is another type of neural network that creates outputs based on sequential information and pattern recognition.

Robotics It is a branch of computer science that is focused on designing and building robots which can imitate human intelligence and actions.


Strong AI It is also an artificial general intelligence that has the ability to perform most of the tasks that a human can do.

Structured data It is a set of data that has easily searchable patterns.

Supervised learning It is a type of machine learning where the output data sets teach it to create better outcomes in the future.


Turing Test It is a test that was created by computer scientist Alan Turing in 1950 to test if machines displayed intelligence that is similar to a human.


Unstructured data It is data that does not have easily searchable patterns.

Unsupervised learning It is a type of machine learning where an algorithm is trained using data that is neither classified nor labeled. This allows the algorithm to behave without guidance.


Weak AI It is alternatively known as narrow AI that has skills to perform only a single or small set of tasks. 

Subscribe to our newsletter

Simple answers to frequently asked questions about AI

Want to learn more about each of these AI terms?

Head over to our Wiki.

Related articles


Responsible AI

Responsible AI is a framework that registers how an organization is addressing the challenges around artificial intelligence (AI) from an ethical and legal perspective.

Read More »

AI platform for the world’s
data-driven companies