Bias

A family of large language models trained to predict the next word in a sequence, then adapted for specific tasks such as drafting or summarizing.

When an AI model generates information that is false, fabricated, or not actually supported by its input, while presenting it with the same confidence as accurate content.

A design approach in which a person reviews, edits, or approves AI-generated output before it is finalized or used, rather than letting the AI operate unsupervised.

A model trained on very large volumes of text so it can predict, generate, and manipulate human language rather than simply search or sort it.

A branch of AI in which a system improves at a task by learning patterns from data rather than following rules a person explicitly programmed.

The type of data or input a system can understand and process, such as text, speech, or images.

A subfield of AI in which a system transforms structured data or inputs into readable, human-like text.

The field of AI focused on how computers understand, interpret, and generate human language rather than structured data or code.

A system of interconnected algorithms, loosely modeled on the brain, that identifies relationships in data and is the structure most modern AI models are built on.

An interactive environment for experimenting with, testing, or fine-tuning an AI model's behavior in a low-stakes setting.

The natural-language instruction a person enters into an AI system to produce a response.

The practice of writing clear, structured instructions that reliably produce useful, accurate output from an AI model.