A – C

Algorithm

A set of step-by-step instructions that tells a computer how to solve a problem.

Artificial Intelligence (AI)

Technology that enables machines to perform tasks that normally require human intelligence.

Agentic AI

AI that can take actions on its own, not just respond to prompts.

API (Application Programming Interface)

A way for different software systems or apps to communicate with each other.

Bias

Unfair or skewed results produced by an AI model due to flawed or unbalanced training data.

Chatbot

An AI program that talks to users through text or voice.

Cloud Computing

Using remote servers (instead of your computer) to store data and run applications.

Computer Vision

AI that can “see” and understand images or videos.

Chain of Thought (CoT)

AI reasoning step-by-step to improve accuracy.

D – F

Data Mining

Extracting useful patterns or information from large datasets.

Deep Learning

A type of AI based on neural networks that learn from large amounts of data — used in image recognition, speech, etc.

Dataset

A collection of data used to train or test AI models.

Diffusion Models

AI models used to generate images or art (e.g., Midjourney, DALL·E).

Embedding

Turning text, images, or data into numbers that represent meaning — used in search and recommendations.

Fine-Tuning

Training an AI model on specific data to improve its performance for a particular task.

Foundation Model

A large, pre-trained AI model (like GPT, Gemini) that can be adapted for different tasks.

G – J

Generative AI

AI that can create new content such as images, text, code, or music.

GPU (Graphics Processing Unit)

A powerful computer chip used to train AI models quickly.

Hallucination

When AI makes up information that isn’t true or correct.

Inference

When a trained AI model uses what it has learned to make predictions or give answers.

Jacobian (rare for beginners)

A mathematical concept used inside some neural network calculations — not typically needed for basic understanding.

K – M

Knowledge Base

A collection of information that an AI uses to answer questions (documents, FAQs, PDFs).

LLM (Large Language Model)

A type of AI that understands and generates human-like text.

Machine Learning (ML)

Teaching computers to learn patterns from data.

Model

A program trained to recognize patterns or make predictions.

Multimodal AI

AI that can understand multiple types of input (text + images + audio + video).

N – P

Neural Network

A computer system inspired by the human brain — made of layers of “nodes.”

NLP (Natural Language Processing)

AI’s ability to understand and generate human language.

Overfitting

When a model learns training data too well but performs poorly on new data.

Parameter

A value inside a model that it learns during training.

Prompt

The instruction you give to an AI system.

Prompt Engineering

Crafting effective prompts to get better AI responses.

Q – S

Quantization

Making AI models smaller and faster by reducing precision — useful for mobile devices.

RAG (Retrieval-Augmented Generation)

A method where AI searches a knowledge source before answering → improves accuracy.

Robotics

Field where AI is used to control physical machines like drones or humanoid robots.

Supervised Learning

Training an AI model using labeled examples.

Synthetic Data

AI-generated data used for training when real data is limited or sensitive.

T – V

Training

The process of teaching a model using data.

Token

A fragment of text (like a word or piece of a word) that AI reads.

Transformer

A type of neural network architecture that powers modern AI models like GPT.

Vector Database

A special database used to store embeddings for fast AI-powered search (e.g., Pinecone).

Vision AI

AI that analyzes and understands images or videos.

W – Z

Weight

A learned value inside a neural network that affects predictions.

Zero-Shot Learning

AI completing tasks it has never been specifically trained for.