Companies Train AI to Act Their Way

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-explorers-foundations-AI-and-the-fine-tune-process-r10a5

1. What is fine-tuning in artificial intelligence?

   1. A method that makes AI forget everything it learned
   2. A way to teach an existing AI to behave in a special way for a specific job
   3. A process that builds a completely new AI from the beginning
   4. A technique that turns computers into physical robots

2. What do companies typically start with before they fine-tune an AI?

   1. A giant base AI that already knows a lot
   2. An AI that only knows how to play games
   3. A small and simple AI for beginners
   4. A blank AI that knows nothing

3. The lesson compares fine-tuning to teaching a smart kid your house rules. What does this comparison show?

   1. Fine-tuning only works on very young AI
   2. Fine-tuning makes AI less intelligent
   3. Fine-tuning makes AI forget everything and start over
   4. Fine-tuning teaches an already-smart AI new ways to behave

4. A company wants an AI that always answers politely to customer questions. What should they do?

   1. Fine-tune a base AI with examples of polite responses
   2. Build a brand new AI from the beginning
   3. Delete all the AI's previous knowledge
   4. Teach the AI to be rude first, then change it

5. What does fine-tuning actually do to a base AI?

   1. It makes the AI unable to learn new things
   2. It polishes and adjusts the AI for a special job
   3. It rebuilds the AI completely from nothing
   4. It erases all the AI's knowledge

6. Which statement about fine-tuning is TRUE?

   1. Fine-tuning only works on computers without internet
   2. Fine-tuning builds AI from scratch
   3. Fine-tuning teaches a base AI to act in a special way
   4. Fine-tuning makes AI forget everything it knows

7. A company wants an AI that writes computer code in Python. What would they do?

   1. Build a new AI that only knows how to speak English
   2. Delete all programming knowledge from the AI
   3. Teach the AI to write poetry instead
   4. Fine-tune a base AI with many examples of Python code

8. What is the main purpose of fine-tuning an AI?

   1. To turn the AI into a physical machine
   2. To break the AI so it cannot be used
   3. To make the AI forget everything it learned
   4. To make the AI behave in a special way for a specific job

9. Why do a coding AI and a chat AI feel different when you use them?

   1. One was fine-tuned for coding and the other for chatting
   2. They were built by completely different companies
   3. Neither one uses fine-tuning
   4. They are actually the same AI

10. What can fine-tuning NOT do?

    1. Build an AI completely from scratch
    2. Help an AI respond in a certain way
    3. Make an AI better at a specific task
    4. Teach an AI new behaviors for a special job

11. A company takes a base AI and gives it thousands of examples of polite conversations. What is this process called?

    1. Fine-tuning
    2. Ignoring
    3. Deleting
    4. Breaking

12. Which of these is an example of fine-tuning?

    1. Teaching a brand new AI its first words
    2. Deleting all files from a computer
    3. Building a computer from parts
    4. Taking a base AI and training it with medical examples to help doctors

13. What kind of examples would you use to fine-tune an AI to write better code?

    1. Examples of good computer code and how it works
    2. Examples of different weather patterns
    3. Examples of people cooking food
    4. Examples of people singing songs

14. When a company fine-tunes an AI, what happens to the knowledge the AI already had?

    1. It stays in the AI and the fine-tuning adds more
    2. It is shared with other computers
    3. It is completely erased
    4. It becomes corrupted and wrong

15. Why do companies use fine-tuning instead of building new AI from scratch?

    1. It is faster and easier because the AI already knows a lot
    2. Fine-tuning is the only method that exists
    3. Building from scratch is cheaper
    4. Building from scratch creates better AI