What is an AI agent?
An AI agent is a language model that can plan steps, use tools and act on your behalf until a goal is met. Learn how it differs from a chatbot.
Scroll, swipe or press ↓ to begin
A chatbot answers a question. An agent works towards a goal.
Give it an outcome, not a question
You describe the result you want, such as “book a meeting with the three people who replied yes”. The agent decides the steps, uses tools and checks its own progress until the job is done.
How a chatbot works
A chatbot takes one message and returns one reply. Every step after that is yours: you read the answer, decide what to do next and ask again.
How an agent works
An agent runs a loop. It looks at the goal, picks the next action, takes it, reads the result and decides again.
You set the goal and the limits
The agent handles the steps in between. Your job moves from doing each step to describing the outcome, the rules and when to stop.
The three parts of every agent
- A model that reads the situation and decides what to do next
- Tools it can call, such as search, email, a spreadsheet or your calendar
- Instructions that set the goal, the rules and when to stop
The model decides
The language model is the brain of the agent. It reads the goal and everything that has happened so far, then chooses the next step.
Tools let it act
Without tools a model can only write text. With tools it can search the web, send an email or add a row to a sheet.
Instructions set the rules
Instructions are plain language. A short, clear brief does most of the work:
Goal: book a 30-minute call with everyone who replied "yes".Use: calendar, email.Rules: only weekdays, 10:00–17:00.Stop: when every invite is sent.