> ## Documentation Index
> Fetch the complete documentation index at: https://stem-docs.intellectualpoint.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Camp wrap-up

> Share what surprised you, hear the facilitator's closing message, and write the last piece of code you will run all week.

## Camp wrap-up

<Info>
  **Time**: 12:15 PM to 12:30 PM
</Info>

### Share your experience

Each student shares:

1. **One thing the robot did this week that surprised you**
2. **One question you still have about AI**

***

### Facilitator close

> "You built something that sees, hears, thinks, and moves. You understand how it works at every layer — from the I2C bus to the API call to the system prompt. Most people who use AI daily have no idea how any of it functions. You do. That puts you ahead of 99% of people talking about AI right now."

***

### Final activity: the goodbye message

Write a goodbye message that the robot speaks aloud. This is the last code you run all week.

```bash theme={null}
sudo python3 -c "
import os
message = input('  Type your goodbye message: ')
safe = message.replace(\"'\", \"'\\\\''\")
os.system(f\"espeak --stdout -s 140 '{safe}' 2>/dev/null | paplay 2>/dev/null\")
print('  Robot said goodbye!')
"
```

Or if you want to get fancy, open Python and write a multi-line farewell:

```python theme={null}
import os

def speak(text):
    safe = text.replace("'", "'\\''")
    os.system(f"espeak --stdout -s 140 '{safe}' 2>/dev/null | paplay 2>/dev/null")

speak("Thank you for an amazing week.")
speak("I learned to see, hear, think, and move.")
speak("I am PiCar X, and I will miss you.")
speak("Goodbye, and keep building cool things!")
```

***

### The week in review

Here is everything you accomplished, day by day:

| Day       | What you built                                                                                      |
| --------- | --------------------------------------------------------------------------------------------------- |
| **Day 1** | Flashed SD card, connected via SSH, calibrated servos, validated all sensors                        |
| **Day 2** | Movement programs, keyboard driving, text-to-speech, line following, color detection, face tracking |
| **Day 3** | LLM concepts, OpenAI API, speech-to-text, voice chatbot, LLM-controlled robot with tool calls       |
| **Day 4** | Vision language models, camera-to-VLM pipeline, vision chatbot, vision-triggered tool calls         |
| **Day 5** | AI ethics, data privacy, robot reporter project, continued learning resources                       |

<Accordion title="Detailed week recap — every skill you learned">
  **Day 1: Setup and Calibration**

  * Flashed Raspberry Pi OS with Pi Imager
  * Connected via SSH over the network
  * Ran the full setup script (robot-hat, vilib, picar-x, espeak, I2S audio)
  * Calibrated steering, pan, and tilt servos
  * Validated ultrasonic, grayscale, servos, motors, camera, battery, and speaker

  **Day 2: Code and Computer Vision**

  * Learned the PiCar-X movement API (forward, backward, steering)
  * Built a movement demo from scratch
  * Built a keyboard-controlled driving program
  * Explored espeak text-to-speech (voices, speeds, pitches)
  * Built a talking driver (TTS + movement combined)
  * Built a line-following program using grayscale sensors
  * Built color detection using OpenCV
  * Built face tracking using Haar cascades

  **Day 3: GenAI and Cloud LLMs**

  * Understood tokens, training, inference, temperature, and context windows
  * Learned what tool calling is and why it matters
  * Connected to the OpenAI API and tested GPT-4o-mini
  * Tested Vosk offline speech-to-text
  * Built a voice chatbot (STT → LLM → TTS)
  * Built an LLM-controlled robot with 10 tools
  * Customized system prompts and wake words

  **Day 4: Vision AI**

  * Understood how VLMs process images alongside text
  * Captured camera frames and sent them to GPT-4o-mini vision
  * Built a vision chatbot that describes what it sees
  * Added `describe_surroundings` as a tool the LLM can call
  * Built the full vision robot: sees, hears, thinks, speaks, and acts

  **Day 5: AI Ethics and Final Project**

  * Discussed bias, hallucination, and AI-generated misinformation
  * Examined surveillance risks using the same tech stack you built
  * Understood data privacy implications of cloud API usage
  * Built the robot reporter: autonomous tour with VLM narration
  * Explored career paths and continued learning resources
</Accordion>

***

### The five layers of your robot

```mermaid theme={null}
flowchart TD
    subgraph layers [What You Built]
        L1["Layer 1: Movement\n(motors, servos, steering)"]
        L2["Layer 2: Sensing\n(ultrasonic, grayscale, camera)"]
        L3["Layer 3: Speech\n(Vosk STT, espeak TTS)"]
        L4["Layer 4: Language Intelligence\n(GPT-4o-mini, tool calls)"]
        L5["Layer 5: Vision\n(camera → VLM)"]
    end

    L1 --> L2 --> L3 --> L4 --> L5
```

Each layer built on the one before it. By the end, your robot had all five — the same architecture used in autonomous vehicles, warehouse robots, and AI research labs.

***

<Check>
  Congratulations on completing the **Introduction to GenAI and LLMs** workshop. You now have hands-on experience with the full stack of modern AI — from soldering connections to system prompts to ethical implications. Keep building.
</Check>