🎯 Objective

Students will explore the evolution of chatbots by analyzing and comparing ELIZA (1966) with modern AI-powered chatbots like ChatGPT. The goal is to understand differences in reasoning, design, and response generation, while critically evaluating AI progress.

🔍 Learning Outcomes

• Understand how ELIZA worked and its limitations.
• Compare ELIZA with modern AI models in terms of reasoning, response quality, and adaptability.
• Reflect on the impact of conversational AI on society.
• Experiment with chatbot development using a simple Python script.

🛠 Workshop Structure

1️⃣ Introduction (15 min) – The Birth of ELIZA

- Brief presentation on ELIZA’s history and purpose.
- How ELIZA’s pattern-matching technique worked.
- Live demo: Try ELIZA online.

💬 Discussion Questions:

• How does ELIZA respond to questions?
• Does it understand meaning, or just manipulate text?
• What kind of reasoning does it use?

2️⃣ Hands-on Activity: Chatbot Comparison (30 min)

💻 Step 1: Interact with ELIZA and ChatGPT

- Students will test ELIZA (via an emulator) and a modern AI chatbot (e.g., ChatGPT, Google Bard).
- They will ask the same set of questions to both and note the differences.

📝 Example Questions to Ask Both Chatbots:

• "I am feeling sad today."
• "What is the capital of France?"
• "Can you explain how a neural network works?"
• "Tell me a joke!"

📊 Step 2: Comparative Analysis

💬 Discussion:

• How do the responses differ?
• Why is ELIZA’s reasoning so limited?
• What enables modern AI to provide more detailed answers?

3️⃣ Building a Simple Chatbot (40 min)

Now, let’s build our own chatbot using Python!

Students will modify a simple rule-based chatbot similar to ELIZA.

import random

responses = {
    "hello": ["Hello! How can I help you?", "Hi there!"],
    "how are you": ["I'm just a program, but I'm doing well!", "I don't have feelings, but thanks for asking!"],
    "sad": ["I'm sorry to hear that. What's on your mind?", "Would you like to talk about it?"],
    "bye": ["Goodbye!", "See you later!"]
}

def simple_chatbot(user_input):
    user_input = user_input.lower()
    for key in responses:
        if key in user_input:
            return random.choice(responses[key])
    return "I'm not sure how to respond to that."

# Chat loop
print("Simple Chatbot: Type 'bye' to exit.")
while True:
    user_input = input("You: ")
    if user_input.lower() == "bye":
        print("Simple Chatbot: Goodbye!")
        break
    print("Simple Chatbot:", simple_chatbot(user_input))
    

🎯 Task for Students:

• Modify the chatbot to recognize more patterns.
• Add new responses to make it more interactive.

4️⃣ Conclusion & Reflection (15 min)

📢 Group Discussion:

• How did ELIZA’s limitations become apparent in your tests?
• What key advancements make today’s AI more powerful?
• What ethical concerns arise with AI chatbots today?

📌 Final Takeaway:

The evolution from ELIZA to modern AI shows how far Natural Language Processing has come, but it also raises questions about AI understanding, ethics, and trust in AI-generated responses.

🎓 Workshop Wrap-up

• ✅ Students submit their modified chatbot code.
• ✅ Fill out a short reflection on their experience.
• ✅ Discuss AI’s future in chatbots and where it might go next.

🔥 Extensions (For Advanced Students)

• Implement keyword detection to improve the chatbot’s responses.
• Use AI APIs (like OpenAI’s API) to connect their chatbot to a more advanced model.
• Explore ethical implications of chatbots replacing human interactions.