The Million Tiny Knobs Inside an AI Brain

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-million-tiny-knobs-r7a5

1. What are the 'tiny knobs' inside an AI called?

   1. They are called neurons and work like human brain cells
   2. They are called servers and store the AI's memory
   3. They are called weights and control how the AI makes decisions
   4. They are called sensors and help the AI see data

2. What happens to these knobs during AI training?

   1. They are replaced with new ones every time
   2. They are turned all the way to maximum
   3. They are locked in place so they don't change
   4. They get nudged slightly this way and that until the AI improves

3. The lesson compares AI training to tuning a piano. What does the piano represent?

   1. A piano player who learns to play songs
   2. A recording device that captures sound
   3. The AI itself, with strings being like the adjustable knobs
   4. A music teacher who trains the AI

4. In the radio analogy, what represents the AI's 'answer'?

   1. The electricity powering the radio
   2. The antenna receiving signals
   3. The clear song or voice coming through the speaker
   4. The radio dial being turned

5. What does the thermostat analogy teach about AI learning?

   1. The AI controls the thermostat automatically
   2. The AI learns from trial and error to find the best setting
   3. Thermostats don't use knobs, they use buttons
   4. The thermostat always picks the coldest temperature

6. In the paper airplane activity, what do the folds represent?

   1. The person's hands folding the paper
   2. The rules for how to fold
   3. The paper the airplane is made of
   4. The knobs or weights that get adjusted during learning

7. What is the goal of adjusting all these knobs during training?

   1. To make the AI faster at math
   2. To make the AI use less electricity
   3. To teach the AI new languages
   4. To get the AI to give correct answers more often

8. The lesson says some AIs have billions of knobs. Why is such a huge number needed?

   1. Too few knobs would crash the computer
   2. More knobs mean the AI can think faster
   3. Billions of knobs make the AI look more impressive
   4. Complex problems need many tiny adjustments to solve them

9. In the bike seat analogy, what does 'one click at a time' represent?

   1. Using one hand to adjust the seat
   2. Making tiny adjustments rather than huge changes
   3. Clicking the seat into place firmly
   4. Putting only one person on the bike

10. What would happen if someone turned all the knobs in an AI to random positions at once?

    1. The AI would likely give wrong answers because it lost its learned patterns
    2. The AI would become smarter because it started fresh
    3. The AI would refuse to work
    4. The AI would work the same as before

11. After training is complete, can an AI's knobs still change?

    1. Yes, continued training can further improve the knobs
    2. No, training ends forever once finished
    3. No, knobs can only change during special updates
    4. Yes, but only humans can change them

12. Which of these is the best example of 'tuning' like an AI does?

    1. A robot repeating the exact same task forever
    2. A book sitting on a shelf waiting to be read
    3. A calculator solving the same math problem repeatedly
    4. A chef adding a pinch of salt, tasting, then adjusting

13. What did you actually do in the paper airplane activity that is similar to AI training?

    1. Folded the paper exactly the same way every time
    2. Made small adjustments one at a time and tested the result
    3. Drew pictures on the plane to make it fly better
    4. Added more paper to make the plane bigger

14. The lesson mentions 'weights' and 'parameters' as key terms. What are they?

    1. Types of AI that learn differently
    2. Different names for the tiny knobs inside an AI
    3. Programming languages used to build AI
    4. Two separate things - weights are big and parameters are small

15. If you wanted an AI to recognize cats in photos, what would training do?

    1. Teach the AI what a cat looks like by showing it pictures
    2. Install a cat识别 program into the AI
    3. Make the AI able to draw cats
    4. Adjust the knobs so the AI correctly identifies cats more often