AI, Machine Learning & GenAI

When someone in a meeting says “let’s use AI for that,” they could mean almost anything. The phrase has become a catch-all, and that vagueness causes real confusion: people argue past each other, budgets get approved for the wrong thing, and expectations drift apart. The good news is that you don’t need a technical background to follow along. A few clear definitions and a sense of how the pieces nest inside each other will let you ask sharp questions and spot when a word is being stretched too far.

The three words, nested

The simplest way to keep these straight is to picture them as boxes inside boxes. Artificial intelligence (AI) is the big outer box: the broad idea of getting computers to do things that normally need human smarts, like recognising a face or recommending a film. Inside it sits machine learning (ML), the most common way we build AI today. Instead of a programmer writing every rule by hand, an ML system learns patterns from lots of examples. Show it thousands of labelled emails and it learns to flag spam, without anyone listing every rule.

Inside the ML box sits the newest and noisiest piece: generative AI, or GenAI. Where older ML mostly classified or predicted things (“is this spam?”, “will this customer leave?”), generative AI creates new content — sentences, summaries, images, code. The engines behind the chatbots you’ve heard of are large language models (LLMs): systems trained on enormous amounts of text that predict the next word, over and over, to produce fluent writing.

Why the distinction matters at work

Mixing these words up leads to muddled decisions. If a colleague says “the AI will just know our customers,” it helps to ask which kind they mean. A traditional ML model trained on your sales history might predict which customers are likely to renew. An LLM, by contrast, knows nothing specific about your company unless you give it that information in the moment — it learned general language from the public internet, not your private data.

If someone promises that “AI” will solve a problem, ask one question: is it learning from our data, or generating text from general training? The answer changes everything.

A quick mental test

Here’s a shortcut. If the task is prediction or sorting from your own historical data — forecasting demand, scoring leads, spotting fraud — you’re usually in classic machine learning territory, and you’ll need good data and a clear target. If the task is producing language or other content — drafting a reply, summarising a report, rewriting a paragraph — you’re in generative AI territory, and the work shifts toward giving clear instructions and checking the output.

What GenAI is not

GenAI is impressive, but it isn’t a database and it isn’t always right. An LLM produces the most plausible-sounding next words, which usually but not always lines up with the truth. It doesn’t look anything up unless connected to a tool that does. Keeping that in mind protects you from over-trusting a confident paragraph. A model can write a beautifully worded answer that is simply wrong, because fluency and accuracy are not the same thing.

Spot it: AI, ML, or GenAI?

Read each situation and decide for yourself, then tap a card to flip it and check your answer.

Sort the tasks

Drag each task into the bucket it belongs to — or tap an item, then tap a bucket. Hit Check placement when you’re done.

How to use it

In your next AI conversation, listen for which box people are standing in. Try phrases like: “Is this a prediction model trained on our data, or a generative tool?” or “Does this need a machine-learning model, or would a copilot drafting text be enough?” or “When you say AI, do you mean an LLM like a chatbot?” Naming the layer — AI, ML, or GenAI — turns a fuzzy ambition into a concrete project everyone can actually evaluate, scope, and budget.