What it is

A neural network is a computing system loosely inspired by the human brain. Very loosely. It's made up of layers of connected nodes (artificial "neurons") that process information by passing signals between each other. Feed it enough examples and it learns to recognise patterns. It's been around since the 1950s, but only became properly useful when computers got fast enough and data got plentiful enough to make the big ones work.

Why it matters for your job

Neural networks are the plumbing behind most of the AI tools making headlines. Every time someone says "AI can now do X," there's usually a neural network under the bonnet. You don't need to understand the wiring diagram any more than you need to understand how your car engine works. But knowing the basics means you won't be bamboozled by vendors selling you magic beans.

What to do about it

Next time someone pitches an "AI-powered" product to your team, ask what it actually does and how it was trained. Neural networks aren't magic. They're pattern matchers that need good data and clear objectives. Understanding that helps you ask better questions and avoid buying rubbish tools with impressive demos.