Chatbot vs. agent: the one difference that matters

"AI agents" and "agentic AI" are the biggest buzzwords in AI right now — so let's cut through the hype with the one distinction that explains everything: a chatbot answers; an agent acts.

A chatbot or assistant responds to your question in a single turn and then stops. You ask, it replies with words — an answer, a plan, a draft — and you take it from there. It's like a consultant who hands you advice.

An AI agent takes a goal and pursues it over multiple steps, using tools to actually get things done — searching the web, running code, clicking around software, calling other apps — and reacting to what happens along the way. It's like an employee who takes the plan and does the job.

"Agentic AI" is just the umbrella term for AI systems that work this way: they chase a goal across many steps instead of one prompt-and-reply, decide which action or tool to use next, and observe the results (an error message, a search result) before deciding what to do next.

A simple example. Ask a chatbot "what's a good flight to Lisbon?" and it describes options. Ask an agent to "book me a good flight to Lisbon under $600" and — if it has the tools and permission — it searches, compares, and could actually make the booking. The chatbot informs; the agent does.

The honest framing for this course. Agents are genuinely exciting and genuinely useful — but there's a lot of hype, and the reality is more limited than the marketing suggests. This short free course gives you the real picture: what agents are, what they truly can and can't do, the risks, and how to use them safely. By the end you'll understand the hottest topic in AI, no technical background needed — and know where to go if you want to build or automate with agents yourself.

The one line to remember: a chatbot answers, an agent acts — it pursues a goal over multiple steps, using tools, reacting as it goes. That single distinction is the key to understanding everything about AI agents.