Understanding AI: Common terms and acronyms made simple

Artificial Intelligence (AI) is on everyone’s lips these days, but it often comes with a flurry of acronyms and technical terms.  Which can make you feel overwhelmed.  Let’s break down some of the most common AI related terms and help you feel more confident in conversations and this fast-moving space. 

What is Artificial Intelligence (AI)?

Artificial Intelligence (AI) is the ability of a computer system to perform tasks that normally require human intelligence, like understanding text, making decisions or spotting patterns in data. 

What is Machine Learning (ML)?

Machine learning (ML) is a subset of artificial intelligence (AI) that focuses on enabling computers to imitate the way humans learn, perform tasks and try and improve their performance and accuracy through experience and exposure to data. 

What is big data?

Big data refers to extremely large datasets that are too complex or massive to be processed by traditional tools. AI and ML thrive on big data; they can identify patterns, trends and insights. 

What is structured data?

Structured data is highly organised information that is stored in a fixed format, like spreadsheets or databases.  Structured data is easy for computers to search, analyse and use in models or reports. 

What are AI models?

An AI model is a program that has been trained using a data set and has recognised patterns and can make certain decisions without further human intervention. Each model might use different data to train on, apply different algorithms to interpret inputs and process an output depending on how they have been programmed. GPT-4, DALL-E, BERT, are examples of models.

What is Natural Language Processing (NLP)?

Natural Language Processing (NLP) is a branch of artificial intelligence (AI) that aims to bridge the gap between human language and computer systems. NLP enables computers to understand, interpret, and generate human language. 

What is Natural Language Understanding (NLU)?

Natural language understanding (NLU) is a subset of natural language processing (NLP) that focuses on enabling computers to comprehend and interpret human language. Unlike NLP the purpose of NLU is specifically concerned with understanding the meaning, context and intent behind the words and phrases that people use. 

What is a Large Language Model (LLM)?

A large language model (LLM) is a type of artificial intelligence (AI) that can process and produce natural language.  It learns from a massive amount of data from various sources, including books, articles, webpages, and images. 

What are Masked Language Models (MLM)?

Masked language models (MLM) are a type of large language model (LLM) that is used to help predict missing words from text in natural language processing (NLP) tasks. MLM captures the meaning of words by looking at the entire context of a sentence, both to the left and right of the missing word to determine what the correct missing word is. 

What is Retrieval-Augmented Generation (RAG)?

Retrieval-Augmented Generation (RAG) is the technique that helps AI models give more accurate and up to date answers by letting them look things up before responding.  It’s a bit like the “phone a friend” option within the game show Who Wants to be a Millionaire, where the contestant can phone a friend to get some advice before answering the question.

What is prompt engineering?

Prompt engineering is the art of crafting effective instructions (called prompts) to get useful or accurate outputs from AI models, especially LLMs. It’s part science and part art, and it's a crucial skill to learn in order to get AI to do what you want. 

What is ChatGPT?

ChatGPT is an AI chatbot that has been developed by OpenAI that uses a Large Language Model (LLM) to generate human-like responses to text input.  It can answer questions, help with writing, explaining complex topics and much more. 

What is Agentic AI?

Agentic AI refers to artificial intelligence that can act independently, make decisions and pursue goals without needing constant human input.  Traditional AI waits for human input, a set of instructions and then responds, where as Agentic AI behaves more like a digital assistant that understands your goals and works proactively to achieve them. 

What is an MCP Server?

A Model Context Protocol (MCP) server is a special kind of server that helps AI models connect to tools, data and services in a standardised way.  AI models are smart, but they don’t know everything, especially when it comes to real time data or even private data.  By using an MCP server you can connect live data, like your calendar to an AI model, to help give it additional context or access to perform tasks. 

What is OpenAI?

OpenAI is both an organisation and a technology.  OpenAI the organisation was founded in 2015 as a research company with a mission to develop safe and beneficial artificial general intelligence (AGI) that can do tasks better than humans. 

OpenAI technology is known for creating AI models such as ChatGPT, GPT-4, DALL-E, Codex and Whisper. 

What is Azure OpenAI?

Azure OpenAI is a service from Microsoft that gives you access to AI models like GPT-4, DALL-E and Codex through the Azure cloud platform.  It combines OpenAI’s cutting edge technology with Azure’s enterprise grade security, compliance and scalability. Think of it as a business ready version of ChatGPT you can plug into your open apps, websites and workflows. 

What is Azure AI Foundry? 

Azure AI Foundry is Microsoft’s all-in-one platform for building, testing and deploying AI applications.  Think of it as a workshop for AI, a place where developers, engineers, data scientists and business teams collaborate and create smart solutions using tools and models. 

What is Responsible AI?

Responsible AI means building and using artificial intelligence in a way that is safe, fair and trustworthy.  It’s about making sure that AI systems are designed to help people and not cause any harm.  AI systems should also follow ethical rules and legal standards. 

Microsoft’s Responsible AI principles and approaches can be found here: https://www.microsoft.com/en-us/ai/principles-and-approach 

What is low-code?

Low-code is a way to build software applications using visual tools that might have drag-and-drop type interfaces, instead of writing lots of complex code.  Low-code’s aim is to make development faster and easier and open to people who aren’t developers. 

Wrapping up

AI might seem full of buzzwords and acronyms, but once you break them down they can become less daunting.  Whether you're experimenting with tools like ChatGPT or building solutions in Azure or even just trying to keep up with conversations, understanding these key terms gives you a solid foundation. 

Have you got a term or acronym you’d like me to explain? Drop it in the comments and I can get it answered for you!