AI Overture - Part 2

As we talk about AI, AI can be categorized into three major types. We might have already heard of Narrow AI, General AI, and Superintelligence.

Types of AI

Narrow AI

Also known as Weak AI, it is designed to perform a specific task or a set of tasks. It operates within a limited context and does not possess general intelligence or self-awareness. For example, Siri, Alexa, and Google Assistant, they are all belong to Narrow AI. They can help auto-execute or assist with certain tasks. The reason why this kind of AI is called Narrow AI is because it cannot generalize its knowledge to other domains or tasks. For example, Siri cannot drive robot to walk and fetch items.

General AI

Also known as Strong AI, it refers to devices that can understand, learn, and apply knowledge across a wide range of tasks, similar to humans.

General AI does not currently exist. The robot in the movie I, Robot is one example. Achieving General AI requires solving complex problems like self-awareness, reasoning, and emotional intelligence. It is a theoretical concept and a long-term goal of AI research.

Super Intelligence

The ultimate goal is to reach Superintelligence. It refers to AI that surpasses human intelligence in all aspects, including creativity, problem-solving, and emotional intelligence.

Superintelligence is purely speculative at this point. It is often discussed in the context of AI safety and ethics. The Skynet from The Terminator is one example of superintelligence in fiction. The development of superintelligence raises significant ethical and existential questions about control, safety, and the future of humanity.

“The development of full artificial intelligence could spell the end of the human race.” - Stephen William Hawking

Narrow AI example: Simple Chatbot

Let us create a simple AI chatbot using Python together. This chatbot will respond to user inputs based on predefined rules. You can install Python on your machine and use Visual Studio Code to edit it and run it.

# Importing the necessary library
import random

# Define responses for the chatbot
responses = {
    "hello": ["Hi there!", "Hello!", "Hey!"],
    "how are you": ["I'm good, thanks!", "Doing well, how about you?", "All good!"],
    "bye": ["Goodbye!", "See you later!", "Bye!"],
    "default": ["I'm not sure how to respond to that.", "Could you rephrase that?", "Interesting!"]
}

# Function to generate a response
def chatbot_response(user_input):
    user_input = user_input.lower()
    for key in responses:
        if key in user_input:
            return random.choice(responses[key])
    return random.choice(responses["default"])

# Main loop for the chatbot
print("Chatbot: Hi! I'm your friendly chatbot. Type 'bye' to exit.")
while True:
    user_input = input("You: ")
    if user_input.lower() == "bye":
        print("Chatbot: Goodbye!")
        break
    response = chatbot_response(user_input)
    print(f"Chatbot: {response}")

How it works

1. The chatbot uses a dictionary (responses) to store predefined responses for specific inputs.

2. The chatbot_response function checks if the user’s input matches any key in the dictionary and returns a random response.

3. If no match is found, it returns a default response.

4. The chatbot runs in a loop until the user types "bye."

As we mentioned in the previous article, it’s rule-based and it cannot improve or adapt over time. I will show how to use more advanced approaches to make it more adaptable and generally intelligent.

While Narrow AI is already a part of our daily lives, General AI and Superintelligence remain in the realm of research and speculation. The coding example demonstrated how to build a simple Narrow AI chatbot using Python, showcasing the practical application of AI in a very easy-to-understand way.

Since we now have a basic concept of AI and AI types, let’s solve the confusion of some AI-related concepts.

AI vs. Machine Learning vs. Deep Learning

Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) are often used interchangeably, but they are distinct concepts with specific relationships. Let me them down.

The major difference is the scope.
AI is the broadest concept, referring to machines or systems that can perform tasks requiring human-like intelligence.
ML is a subset of AI that focuses on enabling machines to learn from data without being explicitly programmed. It uses algorithms to identify patterns and make predictions or decisions.
DL is a subset of ML that uses artificial neural networks with multiple layers (hence "deep") to model complex patterns in data. It is particularly effective for tasks like image recognition(IR), natural language processing (NLP), and speech recognition(SR).

Examples of AI, ML and DL.

AI - A self-driving car
ML - A spam filter that learns to classify emails based on user-labelled data.
DL - Facial recognition systems used in smartphones or social media platforms.

In this article, we explored the fascinating world of AI, breaking down its types - Narrow AI, General AI, and Superintelligence - and their real-world applications. We also clarified the relationships between AI, Machine Learning, and Deep Learning, and even dived into some hands-on coding with a simple chatbot and data visualization using Python.

Stay tuned for the next part of Xin's AI Exploration, where we’ll explore more AI concepts, codes and applications! Enjoy your wisdom journey with AI. &#129395