Types of Artificial Intelligence

# CHAPTER 3

Types of Artificial Intelligence #

1. Introduction #

2. Learning Objectives #

• Define Artificial Narrow Intelligence (ANI).

• Explain the concept of Artificial General Intelligence (AGI).

• Discuss the theoretical concept of Artificial Super Intelligence (ASI).

• Distinguish between Reactive Machines and Limited Memory systems.

3. Beginner-Friendly Explanation #

• Narrow AI (ANI): The Olympic Sprinter. They are the absolute best in the world at running 100 meters in a straight line. But if you ask them to ice skate or play chess, they have no idea what to do.

• General AI (AGI): The Decathlete. They are good at everything. They can run, jump, swim, and play chess just as well as an average human.

• Super AI (ASI): Superman. They are vastly superior to the best human in every single task imaginable.

Currently, every single piece of AI that exists in the world—including ChatGPT and self-driving cars—is an Olympic Sprinter (Narrow AI).

4. Classification by Capabilities #

1. 1. Artificial Narrow Intelligence (ANI): Also known as Weak AI. It is trained to perform a single, specific task. *Examples: Siri, facial recognition, spam filters, chess engines.*

1. 2. Artificial General Intelligence (AGI): Also known as Strong AI. This is a hypothetical machine with the ability to understand, learn, and apply intelligence to *any* intellectual problem, indistinguishable from a human. *AGI does not exist yet.*

1. 3. Artificial Super Intelligence (ASI): A theoretical machine that is infinitely smarter than the brightest human minds in every field, including creativity, general wisdom, and social skills.

5. Classification by Functionalities #

• Reactive Machines: The most basic AI. They do not store memories or past experiences. They look at the current situation and react. *Example: IBM's Deep Blue chess computer. It looks at the board right now and calculates the best move.*

• Limited Memory: These machines can look into the past for a short period to make better decisions. *Example: Self-driving cars. They observe the speed and direction of other cars over the last few seconds to predict where they will be next.*

• Theory of Mind (Theoretical): AI that understands that humans and other entities have their own thoughts, emotions, and intentions, and can interact socially.

• Self-Aware AI (Theoretical): The ultimate AI that has human-like consciousness and is aware of its own existence.

6. Visual Analogies #

• Narrow AI: An NPC (Non-Player Character) that is programmed only to sell you weapons. It can't do anything else.

• General AI: Another human playing the game with you. They can explore, fight, talk, and learn the rules of the game just like you.

• Super AI: The developer of the game who can rewrite the code while you are playing it.

7. Real-World Examples #

• Narrow AI in action: A medical AI can diagnose skin cancer from photos with 99% accuracy. However, if you show that exact same AI a picture of a broken bone, it will likely output gibberish, because it was only trained on skin. It is "Narrow".

8. The Pursuit of AGI #

9. Mini Project #

1. 1. A Roomba vacuum that maps your house. *(Limited Memory)*

1. 2. A machine playing Tic-Tac-Toe. *(Reactive)*

1. 3. A robot that attends college and writes a thesis. *(AGI)*

10. Best Practices #

• Use accurate terminology: Avoid calling current chatbots "sentient" or "conscious." They are highly advanced Limited Memory/Narrow AI systems doing complex pattern matching.

11. Common Mistakes #

• Confusing Generative AI with General AI (AGI): Just because ChatGPT can write poetry, code in Python, and translate French does not mean it is AGI. It is still a Narrow AI—its specific "narrow" task is predicting the next logical word in a sequence of text.

12. Exercises #

1. 1. Explain in your own words why a self-driving car must be a "Limited Memory" system rather than a "Reactive" system.

13. Coding Challenges #

{
  "system": "self_driving_radar",
  "memory_buffer": [
    { "time": "-3s", "car_ahead_speed": 60, "distance": 100 },
    { "time": "-2s", "car_ahead_speed": 55, "distance": 90 },
    { "time": "-1s", "car_ahead_speed": 40, "distance": 70 }
  ],
  "decision": "APPLY_BRAKES"
}

14. MCQs with Answers #

15. Interview Questions #

• Q: Contrast Artificial Narrow Intelligence (ANI) with Artificial General Intelligence (AGI).

• Q: Why do researchers argue that current Large Language Models (like ChatGPT) are not true AGI?

16. FAQs #

17. Summary #

18. Next Chapter Recommendation #