Use AI to Help Someone Else: A Generosity Practice

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-agentic-AI-help-someone-else

1. Which sentence best captures the main idea of 'Use AI to Help Someone Else: A Generosity Practice'?

   1. Agents and chatbots are the same thing in every way
   2. AI is amazing for helping others. Solve their tech problems, draft hard messages, plan events. Generosity made easy.
   3. Agents should always run without limits or oversight
   4. Tools and goals are unnecessary for agent design

2. Which of the following is part of 'Some examples'?

   1. Grandma cannot figure out her email. AI walks her through it (with you helping).
   2. Ignore cost when scaling
   3. Use the most expensive model regardless of fit
   4. Approve all actions automatically

3. Which of the following is part of 'The rule'?

   1. Run unbounded retries on any error
   2. Use AI to make others' lives easier. It is one of the best uses of the tool. And it builds you a reputation as the helpful person.
   3. Never log what the agent did
   4. Approve all actions automatically

4. Which of the following is part of 'You did it!'?

   1. Great work — you just leveled up.
   2. Never log what the agent did
   3. Disable safety checks for speed
   4. Skip every form of evaluation

5. What is 'generosity' in this context?

   1. A way to disable the agent's tools
   2. A core concept covered in Use AI to Help Someone Else: A Generosity Practice
   3. A reason to skip all logging
   4. A trick to bypass approvals

6. What is 'helping others' in this context?

   1. A trick to bypass approvals
   2. A reason to skip all logging
   3. A way to disable the agent's tools
   4. A core concept covered in Use AI to Help Someone Else: A Generosity Practice

7. What is 'AI for good' in this context?

   1. A core concept covered in Use AI to Help Someone Else: A Generosity Practice
   2. A reason to skip all logging
   3. A way to disable the agent's tools
   4. A trick to bypass approvals

8. Before using an agent to help someone else with their personal accounts, what should you do first?

   1. Get explicit consent and use their account, not yours, with the smallest possible scope
   2. Copy their data to your machine
   3. Just dive in to save time
   4. Promise them it will be perfect

9. An agent quietly retries a failed payment 50 times overnight. What design principle was missing?

   1. A larger model
   2. More creative prompting
   3. Bounded retries with human notification on repeated failure
   4. A bigger context window

10. What is the most reliable way to keep an autonomous agent from going off the rails on a long task?

    1. Run it for as many steps as possible without checking in
    2. Disable its tools so it can only think
    3. Set a clear goal, a step budget, and review checkpoints
    4. Trust the model to know when to stop

11. Which budget control most directly prevents runaway costs from an agent loop?

    1. A longer context window
    2. A hard cap on steps, tokens, or dollars per task
    3. A bigger model
    4. A friendly system prompt

12. Why is it dangerous to give an agent access to your email and calendar without scoped permissions?

    1. It will refuse to work
    2. Broad access means a single misstep can send the wrong message or wipe events
    3. Scoped permissions slow the model down
    4. Scopes only matter for paid accounts

13. Before letting an agent take a destructive action, what is the safest default?

    1. Skip approvals if the user trusts the agent
    2. Require explicit human approval for the specific action
    3. Hide the action from any log
    4. Approve once and let the agent repeat forever

14. Which of these is the strongest indicator that an agent workflow is ready to scale?

    1. It worked once for one user
    2. It used the latest model
    3. It runs without any logging
    4. It passes a repeatable eval, has cost in budget, and a rollback plan

15. What should an agent's trace let you do after a run?

    1. Reconstruct each step, decision, and tool call so you can debug or audit
    2. Replace the need for any tests
    3. Make the agent run faster next time automatically
    4. Hide what the agent did from the user