Quiz: Module 1, Lesson 3 - Flow Engineering: The New Frontier of Prompting
Instructions: Answer the following questions based on the lecture notes for M1, L3.
Part 1: Multiple Choice
1. What is the key characteristic that distinguishes an "AI Agent" from a simple "Chatbot"? a) The ability to understand natural language. b) The ability to remember previous parts of the conversation. c) The ability to reason, plan, and use external tools to accomplish a goal. d) The ability to generate human-like text.
2. What is the core idea behind "Chain-of-Thought" (CoT) prompting? a) Instructing the LLM to use as many tools as possible. b) Instructing the LLM to provide the final answer immediately. c) Instructing the LLM to "think step-by-step" and outline its reasoning process. d) Instructing the LLM to ask clarifying questions before answering.
3. In the context of building AI agents, what is the primary purpose of a framework like LangGraph? a) To train new foundation models from scratch. b) To provide a visual interface for chatting with an LLM. c) To sell AI-powered services to customers. d) To define stateful, agentic workflows as a graph of nodes and edges.
Part 2: Short Answer
4. You are building an AI agent to help a user plan a trip. The user's prompt is: "Find me a flight to Paris and a hotel for next week." Briefly explain the 'plan' the agent might form and what 'tools' it might decide to use.
Answer Key
1. c) The ability to reason, plan, and use external tools to accomplish a goal.
2. c) Instructing the LLM to "think step-by-step" and outline its reasoning process.
3. d) To define stateful, agentic workflows as a graph of nodes and edges.
4. (Sample Answer): The agent would first form a plan, breaking the request into sub-tasks. Plan: 1. Find a suitable flight to Paris for the specified dates. 2. Find a hotel in Paris for the same dates. 3. Present the options to the user. To execute this plan, the agent would use tools. For step 1, it would use a flight_search_tool (which could be an API for an airline or travel site). For step 2, it would use a hotel_search_tool (an API for a hotel booking site). It would then use its internal reasoning capability to synthesize the results before presenting them.