Ai Dictionary

Understand the Language of AI

Confused by technical terms? We've created a simple glossaryto help you navigate the AI concepts we use across our work—so you can stayinformed, not overwhelmed.

Language of Modern AI

Get familiar with the most important terms in artificial intelligence—explained simply. Whether you're new to AI or just need a quick reference, this glossary will help you navigate the tools, models, and concepts shaping the future of tech.

A

AI (Artificial Intelligence)

Technology that enables machines to simulate human-like learning, reasoning, and decision-making.

AI Agent

An autonomous system that performs tasks, makes decisions, and adapts to user goals without constant input.

Anthropic

An AI company known for developing the Claude LLM, focused on safety and alignment.

B

Bias

Systematic error in AI results, often caused by unbalanced or non-representative training data.

Business Intelligence (BI)

Use of data analysis tools (often AI-enhanced) to support better business decision-making.

C

ChatGPT

A widely used AI chatbot developed by OpenAI for generating text, answering questions, and more.

Computer Vision

AI that interprets and processes visual inputs like images or video.

Cloud Computing

Using remote servers (e.g. AWS, Azure) to store data and run AI models.

D

DALL·E

An image-generating AI from OpenAI that creates visuals from text prompts.

Data Labeling

Tagging or annotating data so it can be used to train AI models.

Deep Learning

A subfield of machine learning based on neural networks that solves complex problems.

E

Embedding

Numeric vector representation of text, images, or audio that allows AI to compare and process content.

Ethical AI

The practice of building AI systems that are fair, transparent, and aligned with human values.

F

Fine-Tuning

Adapting a pre-trained AI model to perform better on a specific domain or task.

Foundation Model

A large general-purpose model (like GPT or LLaMA) that can be customized for many applications.

G

Generative AI

AI that creates new content such as text, images, code, or audio based on inputs.

GPU (Graphics Processing Unit)

Hardware used to train and run deep learning models efficiently.

H

Hallucination

When an AI generates content that sounds correct but is factually false or made up.

Hugging Face

A popular platform for open-source AI models and tools.

I

Inference

The stage where a trained model is used to generate outputs or predictions.

Integration

Connecting AI systems with tools like CRMs, email, or other software to create seamless workflows.

J / K / L

LangChain

A framework for building AI applications by combining LLMs with tools, APIs, or databases.

LLM (Large Language Model)

AI models trained on massive text datasets to understand and generate human-like language.

M

Model Drift

Decline in model accuracy over time as the data it sees diverges from its training data.

MLOps

The practice of managing and deploying machine learning models in production environments.

Meta LLaMA

A family of open-source LLMs developed by Meta (Facebook).

N

NLP (Natural Language Processing)

The field of AI focused on understanding and generating human language.

Neural Network

The architecture behind deep learning, inspired by the structure of the human brain.

O

OCR (Optical Character Recognition)

AI that reads printed or handwritten text from images or scanned documents.

OpenAI

Leading AI research company behind GPT, ChatGPT, and DALL·E.

P

Prompt Engineering

Crafting effective inputs to guide AI models toward accurate or useful outputs.

Predictive Analytics

Using AI to forecast future trends or behaviors based on past data.

Q / R

RAG (Retrieval-Augmented Generation)

Combines document search with text generation to produce informed AI responses.

S

Scalability

The ability of an AI system to handle more users or data without loss in performance.

Sentiment Analysis

Detecting emotional tone (positive, negative, neutral) in text using AI.

Stable Diffusion

A generative AI model that creates high-quality images from text prompts.

T / U / V

Token

The basic unit of text (word piece or character) that AI models process.

Training

The process of teaching an AI model using large datasets.

Vector Database

Specialized databases optimized for searching vector representations (embeddings) based on similarity.

W / X / Y / Z

Workflow Automation

Using AI to automate repetitive tasks across systems, like email replies or document classification.