AI Glossary & Terms Dictionary
Complete reference guide to artificial intelligence terminology. From fundamentals to advanced concepts - everything you need to understand AI.
π€ AI Fundamentals
Artificial Intelligence (AI)
The simulation of human intelligence by machines, enabling them to learn, reason, and solve problems.
Machine Learning (ML)
A subset of AI that enables systems to learn and improve from experience without being explicitly programmed.
Deep Learning
Neural networks with multiple layers that learn hierarchical representations of data.
AGI (Artificial General Intelligence)
Hypothetical AI with human-like general cognitive abilities and transfer learning capability.
Narrow AI
AI systems designed for specific tasks rather than general intelligence.
Symbolic AI
AI approach using symbols, rules, and logic rather than statistical learning.
π Machine Learning
Supervised Learning
Training ML models on labeled datasets to predict outputs from inputs.
Unsupervised Learning
Finding patterns in unlabeled data without predefined labels or targets.
Reinforcement Learning
Training agents through rewards and penalties based on actions taken.
Neural Network
Computing systems inspired by biological neural networks, consisting of interconnected nodes.
Transformer Model
Architecture using self-attention mechanisms, basis of modern LLMs like GPT and BERT.
Fine-tuning
Adapting pre-trained models to specific tasks or domains through additional training.
Transfer Learning
Applying knowledge from one ML task to improve performance on another related task.
Hyperparameter
Configuration settings that control the learning process of ML models.
π£οΈ NLP & Language
NLP (Natural Language Processing)
Bridging human language and computational understanding through algorithms.
LLM (Large Language Model)
Massive neural networks trained on vast text data for language understanding and generation.
Token
Basic unit of text processing - words, subwords, or characters in LLM context.
Prompt Engineering
Crafting effective inputs to get desired outputs from language models.
Few-shot Learning
LLM capability to learn new tasks from just a few examples in the prompt.
Zero-shot Learning
LLM ability to perform tasks it was never explicitly trained on.
Text Summarization
Automatically condensing longer texts into shorter versions while preserving key information.
Sentiment Analysis
Determining emotional tone, opinions, or attitudes expressed in text.
π€ AI Agents
Agentic AI
AI systems that autonomously plan, execute, and adapt multi-step actions to achieve goals.
AI Agent
Autonomous system that perceives environment, makes decisions, and takes actions.
Tool Use
AI capability to interact with external tools, APIs, and real-world systems.
MCP (Model Context Protocol)
Standard protocol enabling AI models to connect with external data sources and tools.
RAG (Retrieval-Augmented Generation)
Combining information retrieval with LLM generation for more accurate, grounded responses.
AI Memory
Systems allowing AI to retain and recall information across conversations.
AI Planning
Capability to decompose goals into actionable steps and sequence tasks.
Chain-of-Thought Reasoning
LLM technique of showing intermediate reasoning steps for complex problems.
βοΈ Infrastructure & Data
Vector Database
Specialized database storing embeddings for similarity search and semantic queries.
Embedding
Numerical vector representations capturing semantic meaning of text, images, or data.
Synthetic Data
AI-generated data that mimics real data for training when authentic data is scarce.
Model Distillation
Transferring knowledge from large models to smaller, more efficient ones.
Quantization
Reducing model size by using lower precision numbers for weights and activations.
GPU for AI
Graphics processors repurposed for parallel AI/ML computation acceleration.
Inference
Running trained AI models to generate predictions or outputs on new data.
Model Training
Process of teaching ML models by exposing them to data and adjusting parameters.
π οΈ Tools & Platforms
ChatGPT
OpenAI's conversational AI based on GPT models for natural language interaction.
Claude
Anthropic's AI assistant known for constitutional AI and safe, helpful responses.
Gemini
Google's multi-modal AI model family understanding text, code, images, audio, and video.
GitHub Copilot
AI pair programmer by GitHub and OpenAI for code completion and generation.
LangChain
Framework for developing applications powered by language models.
AutoGen
Microsoft framework for building multi-agent AI applications.
Start Learning AI Today
Bookmark this glossary and check back for new terms added weekly.