AI Fundamentals

Explainable AI (XAI) is the set of methods and practices that make an AI system's outputs understandable to people, showing which factors drove a prediction and why a particular decision was reached.

Black box AI describes systems whose internal decision-making cannot be directly inspected or understood, so users see only inputs and outputs without a readable explanation of how one produced the other.

AI bias, also called algorithmic bias, is systematic unfairness in an AI system's outputs that disadvantages particular groups of people, usually originating in training data, problem framing or deployment context rather than in the algorithm itself.

Responsible AI is the practice of designing, building and operating AI systems so they are fair, transparent, safe, private and accountable, turning broad ethical principles into concrete engineering and governance practices.

AI benchmarking is the practice of measuring model performance on standardised tests so different systems can be compared, using public benchmarks, human preference rankings and internal evaluation sets.

AI decision-making is the use of artificial intelligence to choose, recommend or support an action based on data, ranging from suggesting options to a person to making decisions automatically without human review.

AI recognition is the ability of an AI system to identify what something is from raw input such as images, audio, video or text, for example recognising a face, a spoken word, a product on a shelf or handwriting on a form.

AI classification is the task of sorting input into predefined categories, such as spam or not spam, positive or negative review, or fraud or genuine payment. It is one of the most common and practical uses of AI in business.

AI prediction is the use of a trained model to estimate an unknown or future value based on patterns in past data, such as next month's sales, the chance a customer will leave or the likely delivery time of an order.

AI reasoning is the ability of an AI system to work through a problem in logical steps, combining information to reach a conclusion, solve a task or make a plan, rather than simply matching an input to a familiar pattern.

AI inference is the stage where a trained AI model is put to work: it receives new input and produces an output, such as a prediction, a label or generated text. Every time you use an AI app, inference is happening.

AI automation is the use of artificial intelligence to complete tasks with little or no human effort, especially tasks that involve reading unstructured information, making judgements or predicting outcomes, which traditional rule-based automation cannot handle.

An AI workflow is a series of connected steps in which AI performs one or more tasks, such as reading, classifying, summarising or deciding, to move work from a starting trigger to a finished outcome, often with people reviewing key steps.

An AI application is a piece of software that uses artificial intelligence to help people complete a task, such as writing, searching, translating, analysing data or answering questions. It is the part of AI most people actually see and use.

An AI algorithm is a set of step-by-step instructions that tells a computer how to learn from data or solve a problem. In machine learning, the algorithm is the method, and the trained model is what it produces.