What is an AI agent?
A chatbot is like a friend you ask questions. An agent is like a helper who can also go and do small jobs for you: look something up, sort a list, check a calendar, then come back and tell you what they did. The more a helper can do on their own, the more carefully you have to watch them.
- Anthropic's engineering article "Building effective agents" (December 19, 2024) calls these "agentic systems" and splits them in two:
- Workflows: the AI and its tools follow steps that you planned in advance.
- Agents: the AI decides its own steps and which tools to use, until the task is done.
- The article's main advice is to start with the simplest solution and add complexity only when it clearly makes results better. Many tasks need only one well-written prompt.
- Agents can cost more and mistakes can add up, so they need testing and guardrails.
| Shape | What it is | Example |
|---|---|---|
| Single prompt | One question, one answer | "Summarize this email." |
| Workflow | Fixed steps you designed | Summarize, then draft a reply, then check tone |
| Agent | The AI picks steps and tools | "Find which orders are late and draft apology emails." |
- In this course you build the middle of that table: an assistant with clear instructions and your own files. It is "agent-like" and a safe place to learn. Real agents with tools come in course 1.4 and later.
- Write down three jobs you would like help with this week, one per line.
- Next to each job, write "single prompt" if one good request could do it.
- Write "workflow" if the steps are always the same (for example summarize, then draft, then check tone).
- Write "agent" only if you cannot predict the steps or the tools needed.
- Look at your list. Most jobs will be single prompts or workflows. That is normal and good.
- Check: each job has a label, and you can say in one sentence why you chose it.
Videos by other creators on YouTube, for reference. They are not made by or affiliated with The Future Market.
Read the video script
[Host] "People say 'AI agent' all the time. Here is what it means."
[Text on screen: Chatbot answers. Agent acts.] "A chatbot answers your question. An agent can take steps and use tools, then report back."
[Screen: Anthropic's 'Building effective agents' article] "Anthropic's advice: start with the simplest thing that works."
[Host] "One good prompt first. A workflow next. A full agent only when you really need it."
List three jobs and label each as single prompt, workflow or agent. Keep the list; you will pick one job from it in Lesson 2.
ClaudeStuck? Copy a ready-made prompt for Claude
Copy this, open Claude, paste it, and Claude will walk you through the lesson step by step.
For a grown-up: Claude is for adults (18+). Copy this, paste it into Claude, and use the answer to explain the lesson to your child.
I'm learning "Agents First Steps: what an AI agent is and a simple first one" on The Future Market. Lesson 1: "What is an AI agent?". What the lesson teaches: - Anthropic's engineering article "Building effective agents" (December 19, 2024) calls these "agentic systems" and splits them in two: - Workflows: the AI and its tools follow steps that you planned in advance. - Agents: the AI decides its own steps and which tools to use, until the task is done. - The article's main advice is to start with the simplest solution and add complexity only when it clearly makes results better. Many tasks need only one well-written prompt. - Agents can cost more and mistakes can add up, so they need testing and guardrails. - Shape - What it is - Example - Single prompt - One question, one answer - "Summarize this email." The exercise I'm working on: List three jobs and label each as single prompt, workflow or agent. Keep the list; you will pick one job from it in Lesson 2. I'm stuck. Please: 1. Explain the main idea in plain, simple words with one everyday example. 2. Walk me through the exercise one small step at a time. Ask me what I've done so far before giving the next step. 3. Don't just give me the final answer. Help me do it myself, and check my work when I share it.
I'm a parent helping my child (about 8 years old) with a lesson called "What is an AI agent?" from the course "Agents First Steps: what an AI agent is and a simple first one". Here's what the lesson says: A chatbot is like a friend you ask questions. An agent is like a helper who can also go and do small jobs for you: look something up, sort a list, check a calendar, then come back and tell you what they did. The more a helper can do on their own, the more carefully you have to watch them. The activity: List three jobs and label each as single prompt, workflow or agent. Keep the list; you will pick one job from it in Lesson 2. Please explain this so an 8-year-old understands: short sentences, a fun everyday example, and no hard words. Then give me 3 simple questions I can ask my child to check they understood, and a way to do the activity together, one tiny step at a time.
Sources
- Anthropic, Building effective agents: www.anthropic.com/engineering/building-effective-agents
Pick one small, safe job
This lesson is part of full access: every course and every level for one payment.
Write instructions a new helper can follow
This lesson is part of full access: every course and every level for one payment.
Build it in a Claude Project
This lesson is part of full access: every course and every level for one payment.
Test it on real examples
This lesson is part of full access: every course and every level for one payment.
Safety limits before any tool access
This lesson is part of full access: every course and every level for one payment.
Use it, improve it, know when to level up
This lesson is part of full access: every course and every level for one payment.
Final project
This lesson is part of full access: every course and every level for one payment.
Quiz
What is the difference between a workflow and an agent?
A workflow follows steps you planned; an agent decides its own steps and tools.
What does Anthropic advise to do first?
Start with the simplest solution and add complexity only when it clearly improves results.
What are two traits of a good first job for an assistant?
A clear purpose and a low risk if it is wrong; the output is easy to judge.
What is the friend test for instructions?
If a person with little context would be confused by your note, the AI will be too.
How many examples does Anthropic suggest including for best results?
3 to 5.
What do Claude Projects hold?
Their own chat histories, a knowledge base of uploaded files and custom instructions.
What three kinds of inputs belong in a test sheet?
Normal, tricky and nonsense.
Name the three tool permission settings for a connector.
Always allow, Needs approval and Blocked.
What is prompt injection?
Hidden adversarial instructions inside content an AI processes, such as a web page.
What is a safe first rule for any assistant that can reach an app?
Read-only first; a person approves anything that sends, deletes, pays or posts.
Final exam
This lesson is part of full access: every course and every level for one payment.
Certificate
This lesson is part of full access: every course and every level for one payment.