What an 'AI Agent' Actually Is (and How It's Different From a Chatbot)

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-foundations-ai-agent-vs-chatbot-r10a10-teen

1. What is the primary distinction between a chatbot and an AI agent?

   1. A chatbot answers questions, while an AI agent takes actions to complete tasks
   2. A chatbot can browse the web but an agent cannot
   3. A chatbot uses more advanced AI models than an agent does
   4. A chatbot is always reliable while agents frequently fail

2. Which of the following is an example of an AI agent that functions as an autonomous software engineer?

   1. Anthropic's Computer Use
   2. OpenAI Operator
   3. Google Gemini
   4. Cognition's Devin

3. When an AI agent works on a complex goal, what does it typically do internally?

   1. Process the entire request at once without any intermediate steps
   2. Execute a single command and wait for user input
   3. Send the task to a human worker and report the result
   4. Break the goal into steps, execute them in a loop, and adjust as needed

4. What does 'tool use' mean in the context of AI agents?

   1. The agent can select and use external capabilities like web browsers, code interpreters, or other AI systems
   2. The agent can physically pick up objects
   3. The agent has access to a larger language model than chatbots
   4. The agent chooses which user questions to answer

5. Why should users supervise AI agents rather than trust them completely?

   1. Supervision is only needed for chatbots, not agents
   2. Agents cannot access the internet without supervision
   3. Agents are still unreliable on complex tasks and can make mistakes
   4. Agents are perfect but need human approval for legal reasons

6. Which capability is typical of an AI agent but not of a basic chatbot?

   1. Responding to text questions with relevant answers
   2. Translating text between languages
   3. Generating creative writing like poems or stories
   4. Clicking buttons and filling out forms on websites

7. What is the 'agent loop' that AI agents perform?

   1. A method where the agent continuously generates text in a loop
   2. A security feature that prevents agents from making errors
   3. A repeated cycle of receiving a goal, planning steps, executing them, and checking results
   4. A way for agents to communicate only with other agents

8. If an AI agent is asked to book a restaurant reservation, what specific actions would it take?

   1. Open a browser, navigate to OpenTable, fill out the reservation form, and confirm
   2. Send an email to the restaurant on the user's behalf
   3. Refuse because booking is not an AI capability
   4. Simply tell the user to call the restaurant

9. What is Anthropic's Computer Use API designed to do?

   1. Create virtual AI characters for video games
   2. Generate computer hardware specifications
   3. Translate between programming languages automatically
   4. Allow developers to build agents that control a virtual desktop environment

10. Why are AI agents described as representing the 'next step' in using AI?

    1. They use more data than previous AI systems
    2. They are cheaper to operate than chatbots
    3. They were invented before chatbots
    4. They can take autonomous actions rather than just responding with text

11. What does it mean when an AI agent has 'autonomy'?

    1. The agent can make decisions and take actions without step-by-step human guidance
    2. The agent can learn entirely on its own without any data
    3. The agent works without any internet connection
    4. The agent is owned by a single company and cannot be shared

12. What percentage of success rate do agents currently have on complex tasks, as mentioned in the lesson?

    1. Approximately 90%
    2. Approximately 70%
    3. Approximately 50%
    4. Approximately 30%

13. What happens when an AI agent encounters an error during task execution?

    1. It typically continues its loop, adjusts its approach, and tries again
    2. It immediately stops and reports failure to the user
    3. It restarts the entire task from the beginning without learning
    4. It asks the user to manually fix the error before continuing

14. Which of the following would NOT be a typical characteristic of an AI agent?

    1. Calling other AI systems as tools
    2. Generating a haiku poem on command
    3. Browsing the web to find information
    4. Writing and executing computer code

15. Why might someone choose to use an AI agent instead of a traditional chatbot for data analysis?

    1. Agents are always free to use
    2. Agents can guarantee 100% accurate results
    3. Agents can only analyze text, not numbers
    4. Agents can write and run code to process files like CSVs automatically