Glossary of AI Terms
Welcome to the glossary for "Prompting 101"! This section provides clear, concise definitions for the key Artificial Intelligence (AI) and Generative AI terms you've encountered throughout the course. Understanding these terms will help you navigate the rapidly evolving world of AI with confidence.
A
AI (Artificial Intelligence): The broader field of computer science dedicated to creating machines that can perform tasks typically requiring human intelligence, such as learning, problem-solving, decision-making, and understanding language.
AI Agent: An autonomous AI system that can understand a high-level goal, create a plan, and then take actions across multiple applications and systems to achieve that goal, often without needing step-by-step human guidance.
AI Assistant: A software program that uses AI to help users with various tasks, often through natural language interaction (e.g., drafting emails, summarizing text, answering questions).
Attention Mechanism: A key component within Transformer models that allows the AI to weigh the importance of different words in a sequence, helping it understand context and generate more coherent responses.
B
Bias (in AI): Prejudices or unfair tendencies in an AI model's output, often inherited from the biases present in its training data.
Black Box Problem: The phenomenon where the internal workings and decision-making processes of complex AI models are so intricate that even their creators cannot fully explain why a specific output was generated.
C
Chain-of-Thought Prompting: An advanced prompting technique where you instruct the AI to "think step-by-step" or "show its reasoning" before providing a final answer, improving accuracy for complex problems.
Co-Pilot (Analogy): A metaphor used to describe AI's role as a helpful assistant that works alongside a human user, who remains in control and responsible for the final outcome.
Copyright (in AI context): The legal rights concerning the ownership and use of creative works. In the U.S., purely AI-generated content generally cannot be copyrighted without significant human creative input.
D
Democratization of AI: The trend of making AI tools and capabilities accessible and usable by a wider audience, integrating them into everyday software and platforms.
F
Few-Shot Prompting: An advanced prompting technique where you provide the AI with a few examples (the "shots") of the desired input-output pattern before giving it your actual request, allowing it to learn the pattern.
G
Generative AI: A type of artificial intelligence that can create new, original content (such as text, images, audio, or code) that didn't exist before, based on patterns learned from vast amounts of existing data.
H
Hallucinations (in AI): When an AI generates responses that are fluent and confident but factually incorrect, nonsensical, or fabricated.
L
LLM (Large Language Model): A type of AI model specifically designed to understand, generate, and process human language, trained on massive datasets of text.
M
Multimodal AI: AI systems that can understand, process, and generate content across different types of data (modalities) simultaneously, such as text, images, audio, and video.
P
Prompt: The instruction, question, or input given to an AI model to guide its generation of content.
R
RAG (Retrieval-Augmented Generation): An AI architecture where a language model retrieves information from an external knowledge base before generating a response, ensuring answers are grounded in facts and reducing hallucinations.
Role-Playing (Prompting): An advanced prompting technique where you instruct the AI to adopt a specific persona, role, or character (e.g., "Act as a seasoned travel agent") to influence its tone, style, and expertise.
T
Tokens: The smaller pieces (words, parts of words, characters, punctuation) into which AI models break down text for processing.
Training Data: The massive datasets of information (text, images, audio, etc.) that an AI model "learns" from during its development, which shapes its knowledge and capabilities.
Transformer Architecture: A neural network architecture that revolutionized AI's ability to process sequential data like language, enabling models to "pay attention" to different parts of the input simultaneously.