In this lesson, we'll learn how agentic AI systems think, reason, and plan. You can think of this process as the "brain" inside the agent that helps it make smart decisions on its own.

When you plan your day, you probably think through steps like this:

• Decide on a goal, like "Finish my homework."

• Think of tasks you need to complete, such as:

  • Reading a chapter

  • Taking notes

  • Solving practice questions

• Put tasks in order, starting with the most important or necessary ones first.

• Adjust your plan if something changes, like if your internet stops working.

Similarly, an AI agent uses reasoning and planning to complete tasks and reach goals. Let's explore how this works step-by-step.

Step 1 – Setting Clear Goals

First, an agent needs a clearly defined goal. A clear goal gives direction and focus. Examples:

• "Plan a movie night."

• "Book a doctor's appointment."

• "Organize a workspace."

Step 2 – Reasoning and Decision-Making

Once an agent knows its goal, it begins to reason. Reasoning simply means deciding how to reach the goal. It involves:

• Thinking about available information.

• Understanding the situation clearly.

• Considering different ways to accomplish tasks.

For example, if the goal is "plan dinner," reasoning might involve:

• Checking what's in the fridge.

• Finding recipes based on available ingredients.

• Considering how much time is available.

Step 3 – Creating a Plan (Step-by-Step Actions)

After reasoning, the agent creates a plan. A plan breaks down tasks into manageable steps, just like a checklist:

• Step 1: Check available ingredients.

• Step 2: Choose an appropriate recipe.

• Step 3: Prepare and cook ingredients.

• Step 4: Set the table and serve food.

This step-by-step plan helps the agent stay organized and clear about what needs to happen next.

Step 4 – Adapting Plans on the Fly

A great strength of agentic AI is flexibility. If something changes, the agent can quickly adjust the plan. Suppose the agent planned to make spaghetti but discovers there's no sauce:

• The agent quickly reasons again and updates the plan.

• It picks a different recipe or orders food instead.

This flexibility makes agentic AI practical and reliable.

Simple Example: Planning a Trip

Let’s use a simple example to put it all together:

Goal: "Plan a weekend trip."

• Reasoning:

  • Identify available options (e.g., beach, mountains, city).

  • Check weather forecasts, travel costs, accommodation availability.

• Planning:

  • Step 1: Pick a destination based on budget and preferences.

  • Step 2: Book transportation.

  • Step 3: Book a place to stay.

  • Step 4: Plan daily activities.

• Adaptation:

  • If hotels at the chosen location are expensive or full, the agent quickly adjusts to another location or accommodation type.

What's Inside an AI Agent's "Brain"?

When we talk about an agent’s "brain," we mean the logic or process that enables reasoning and planning. It typically includes:

• Memory: Remembering past experiences and useful data.

• Logic/Decision Rules: Clear steps or rules for making decisions.

• Adaptability: Flexibility to quickly adjust plans when conditions change.

Small Exercise to Practice

Try making your own reasoning and planning steps. Imagine your agent has this goal:
"Prepare for a friend’s visit."

• Clearly define tasks (cleaning, shopping, arranging guest space).

• Organize these tasks logically.

• Think of how you'd adjust if your friend suddenly arrives earlier than planned.

This helps you appreciate how reasoning and planning work practically in AI.