📑 Contents

Chapter 7.3: Artificial Intelligence (AI)

9618 AS Computer Science

📚 Learning Objectives
🌟 Did You Know?

Artificial Intelligence is playing a key role in furthering research in medicine, where it is used in expert systems to draw connections between illnesses to form diagnoses. AI is becoming essential for today's time because it can solve complex problems efficiently in multiple industries!

Artificial Intelligence LEARN acquire data DECIDE make choices ACT autonomously REASON draw inferences

1. What is Artificial Intelligence?

Artificial Intelligence (AI) is the ability of a computer to replicate human intelligence, cognitive ability, and grasp abstract concepts. AI is a machine or application which carries out a task that requires some degree of intelligence when carried out by a human being.

📖 Definition of AI

AI is a machine that can simulate intelligent behaviours similar to that of a human. It can:

1.1 Tasks That Require AI

These are tasks that would require intelligence when performed by humans:

📝 Examples of AI Tasks

1.2 Types of AI

Type Description Example
Weak AI (Narrow AI) Designed to perform a specific task or set of tasks Voice assistants, chess programs, recommendation systems
Strong AI (AGI) Designed to perform any intellectual task that a human can do Theoretical - not yet achieved
💡 Exam Tip

Remember: Weak AI = Specific tasks (what we have today), while Strong AI = Human-level intelligence (still theoretical). Most AI applications we use daily are examples of Weak AI!

WEAK AI (Narrow AI) Specific tasks only ✓ Available now Evolution STRONG AI (AGI) Any human task ✗ Still theoretical

2. Characteristics of AI

AI systems share three common characteristics that enable them to function intelligently:

Characteristic Description Function
Collection of Data AI systems require large amounts of data to perform tasks Data is processed using rules and algorithms to make decisions and predictions
Rules for Using Data AI systems use logical reasoning to evaluate information Algorithms enable the system to make decisions based on data
Ability to Reason AI can change its own rules and data based on experience Enables learning and improvement over time
DATA Collection & Storage RULES Algorithms & Logic REASON Adaptive Learning Feedback Loop

2.1 Advantages and Disadvantages of AI

Advantages Disadvantages
Increased efficiency - Tasks completed faster Job losses - Automation replaces human workers
Increased accuracy - Reduces human error Biased decision making - Reflects training data bias
Scalability - Can handle massive workloads Ethical concerns - Privacy, accountability issues
24/7 availability - Works continuously High costs - Development and maintenance
Safety - Can work in dangerous environments Dependency - Over-reliance on AI systems

3. Applications of AI

3.1 Expert Systems

Expert systems, also known as knowledge-based systems, replicate the knowledge and experience an expert in a particular subject would have. They are used in medicine to draw connections between illnesses to form diagnoses.

📖 Components of Expert Systems
📝 How Expert Systems Work
  1. Expert knowledge is gathered and coded into rules
  2. Rules stored in the knowledge base (IF-THEN rules)
  3. User inputs symptoms or data
  4. Inference engine matches input against rules
  5. System provides diagnosis or recommendation
Example: A medical diagnosis expert system contains knowledge from many doctors. When a patient's symptoms are entered, the system can suggest possible conditions that even a single doctor might not have considered. However, if faced with a new situation not covered by its rules, it cannot attempt a creative approach like a human doctor.

3.2 Neural Networks

Neural networks are one of the most common uses of AI. They can apply learned knowledge to new data sets, similar to how humans learn. This is used in pattern detection and identifying financial fraud.

USER Input KNOWLEDGE BASE • Facts • Rules • IF-THEN logic INFERENCE ENGINE • Matching • Reasoning • Conclusions OUTPUT Diagnosis

3. Applications of AI (Continued)

3.3 Machine Learning

Machine learning takes place when a system that has a task to perform is seen to improve its performance as it gains experience. The AI system has access to 'experience' in the form of a massive set of data, and learns from this data using algorithms.

📝 Examples of Machine Learning

1. E-commerce Recommendations:

2. Spam Email Detection:

3.4 Perception (Robotics)

Robots have been used in manufacturing processes for repetitive tasks. However, traditional robots continue operating regardless of unexpected events. Autonomous robots are fitted with sensors to enable appropriate action based on sensor information.

📖 Autonomous Robot Features
Example: Driverless Cars
Driverless cars use AI to navigate roads safely. They can park themselves in vacant spaces, detect obstacles, and make driving decisions without human input. Sensors continuously monitor the environment, while AI processes this data to control steering, acceleration, and braking.

3.5 Linguistics (Voice Recognition)

Voice recognition and voice synthesis techniques are already developed and in use. Smart home devices use these technologies but raise questions about privacy, as they must be constantly switched on to function.

⚠️ Privacy Concerns with Voice AI
TRAINING DATA Massive sets ALGORITHM Pattern Recognition TRAINED MODEL Learning OUTPUT / ACTION Prediction Feedback for Improvement

4. Impact of AI on Society

AI can dramatically improve the efficiencies of our workplaces and augment the work humans can do. When AI takes over repetitive or dangerous tasks, it frees up the human workforce for tasks involving creativity and empathy.

4.1 Workforce Changes

📖 Employment Structure Changes

4.2 Education & Accessibility

⚠️ Digital Divide Concerns

4.3 Healthcare Applications

🌟 AI in Healthcare
❌ Key Accountability Questions
BENEFITS Efficiency • Accuracy CONCERNS Jobs • Ethics • Privacy SOCIAL IMPACT

5. Impact of AI on Economy

5.1 Employment & Industry

📖 Economic Restructuring

5.2 Business & Innovation

📝 AI as Catalyst for Economic Growth
⚠️ Competition Concerns

5.3 Market Dynamics & Inequality

Economic Challenge Impact
Wealth Concentration Large tech companies control key AI tools and data
Monopolistic Advantages Reduced market competition
Income Inequality Deepening gap between AI-skilled and traditional workers
💡 Policy Considerations

Policymakers must consider new economic models and regulations to ensure fair access to AI technologies and prevent deepening income inequality. The goal is inclusive growth that benefits all of society.

6. Impact of AI on Environment

6.1 Energy Consumption

⚠️ AI's Energy Footprint
📝 Solutions for Energy Efficiency

6.2 Climate Modelling & Sustainability

🌟 AI for Environmental Protection

6.3 E-Waste and Hardware

Environmental Issue Description
Hardware Demand AI drives demand for specialised hardware (GPUs, TPUs)
Shortened Lifespan Rapid advancements make devices obsolete quickly
E-Waste Increase More electronic waste, pressuring recycling systems
Raw Materials Limited supply of materials needed for AI hardware
Robots and Environmental Impact: Robots can work in environments dangerous for humans, making them very useful. However, the environmental impact of robot manufacture and disposal is significant. They require materials for construction (limited supply) and eventually end up as e-waste, harming the environment.
ENVIRONMENT Balance ENERGY Consumption E-WASTE Hardware CLIMATE Modelling SMART Grids − NEGATIVE + POSITIVE

7. Asimov's Laws of Robotics

People often associate AI with science fiction, fantasy, and robots. The science fiction author Isaac Asimov produced his own three laws of robotics:

⚠️ The Three Laws of Robotics

First Law:

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

Second Law:

A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

Third Law:

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

7.1 Ethical Considerations

❌ Key Ethical Questions

7.2 AI Beyond Robotics

AI goes way beyond robotics. It covers an ever-increasing number of areas:

📝 Diverse Applications of AI
LAW 1 Human Safety LAW 2 Obey Humans LAW 3 Self-Preservation ↑ Priority

8. Key Takeaways

📌 Summary Points

What is AI?

Applications

Impacts

Ethics

9. Glossary of Key Terms

📖 AI Terminology

Artificial Intelligence (AI) → A machine that can simulate intelligent behaviours similar to that of a human

Weak AI (Narrow AI) → AI designed to perform a specific task or set of tasks

Strong AI (AGI) → AI designed to perform any intellectual task that a human can do

Expert System → A knowledge-based system that replicates expert knowledge in a specific domain

Knowledge Base → A set of facts and rules used by expert systems

Inference Engine → The component that interrogates the knowledge base to draw conclusions

Machine Learning → When a system improves its performance as it gains experience

Neural Network → AI that can apply learned knowledge to new datasets, similar to human learning

Autonomous Robot → A robot fitted with sensors to take appropriate action without human intervention

Digital Divide → The gap between those who have access to technology and those who do not

Algorithm → A set of rules that enables the AI system to make decisions and predictions

10. Exam-Style Questions

1. Describe what is meant by Artificial Intelligence (AI). [3 marks]

Answer:

  • AI is a machine or application that carries out tasks requiring intelligence when done by humans
  • It can simulate intelligent behaviours similar to humans
  • AI can learn (acquire new information), decide (make choices), and act autonomously
  • It can process language, recognise faces, and operate machinery
  • AI uses data and algorithms to make decisions and predictions
2. Explain the difference between Weak AI and Strong AI. [4 marks]

Answer:

  • Weak AI (Narrow AI) is designed to perform a specific task or set of tasks
  • Examples include voice assistants, chess programs, recommendation systems
  • Strong AI (AGI) is designed to perform any intellectual task that a human can do
  • Strong AI is still theoretical and has not been achieved yet
  • Weak AI exists today and is used in many applications
  • Strong AI would require human-level consciousness and reasoning
3. Describe the three characteristics of AI systems. [3 marks]

Answer:

  • Collection of data: AI systems require large amounts of data to perform tasks
  • Rules for using data: AI uses logical reasoning and algorithms to evaluate information
  • Ability to reason: AI can change its own rules and data based on experience
  • Data is processed using algorithms to make decisions and predictions
  • Reasoning enables learning and improvement over time
4. A CCTV system uses Artificial Intelligence (AI) to identify the presence of a person in the house and to track their movements. Describe how AI is used in this system. [4 marks]

Answer:

  • The system has been trained using images of humans to recognise patterns
  • AI uses pattern recognition to identify human shapes and movements
  • The system processes video frames in real-time
  • It compares detected patterns against its knowledge base of human characteristics
  • When a match is found, it tracks movement by following the detected person
  • The system may alert homeowners or security services
  • Machine learning allows the system to improve accuracy over time
5. Aisha manages a team of software developers. The team are developing a computer game where the user plays a board game (such as chess) against the computer. Describe how the computer would use AI to play the board game. [4 marks]

Answer:

  • The rules and past moves of the game will be stored
  • The AI program is trained by playing many times
  • AI will look ahead at possible moves
  • AI analyses patterns of past choices
  • AI chooses the move most likely to be successful
  • Computer could learn from previous mistakes
  • It stores positive/negative results of choices and changes future decisions

10. Exam-Style Questions (Continued)

6. Explain how an expert system works in medical diagnosis. [5 marks]

Answer:

  • Expert knowledge is gathered from medical experts and coded into rules
  • Rules stored in a knowledge base (IF-THEN statements)
  • Patient symptoms are input by the user
  • The inference engine matches input against stored rules
  • System provides diagnosis or recommendations
  • Contains more knowledge than a single doctor could have
  • Cannot handle new situations not covered by its rules (unlike humans)
7. Discuss the social impacts of AI on the workforce. [6 marks]

Answer:

  • Job displacement: Automation replaces roles relying on routine/manual work
  • New job creation: AI-related roles emerge (data scientists, ML engineers)
  • Reskilling needed: Workers must adapt to AI-driven transformations
  • Digital divide: Those with better technology access benefit more
  • Work-life balance: More leisure time possible but job security concerns
  • Skills gap: Demand for AI knowledge increases
  • Some argue technology makes rich richer and poor poorer
  • Human-AI collaboration becomes increasingly important
8. Describe the environmental impacts of AI, both positive and negative. [6 marks]

Answer:

Negative Impacts:

  • Large AI models require vast computing power
  • High electricity usage leads to carbon emissions
  • E-waste increase from specialised hardware (GPUs, TPUs)
  • Rapid obsolescence of AI hardware
  • Limited supply of raw materials for AI hardware

Positive Impacts:

  • Climate modelling for weather prediction and environmental analysis
  • Energy optimisation in smart grids and buildings
  • Sustainable agriculture through data analysis
  • Robots can work in environments dangerous for humans
9. Explain what is meant by machine learning and give two examples. [5 marks]

Answer:

  • Machine learning takes place when a system improves its performance as it gains experience
  • AI has access to 'experience' in the form of a massive dataset
  • System learns from data using algorithms
  • Example 1 - E-commerce: Actions of users stored, AI identifies products to advertise, sales increase shows learning
  • Example 2 - Spam detection: Program investigates emails, decides if spam, refuses entry to inbox, improves over time
  • Additional example: Voice recognition improving accuracy based on usage
  • Learning can be supervised, unsupervised, or reinforced
10. Discuss the ethical concerns associated with AI development. [6 marks]

Answer:

  • Accountability: Who is responsible when AI makes wrong decisions?
  • Privacy: Voice assistants must be "always listening" - who has access to data?
  • Bias: AI can perpetuate human biases from training data
  • Job displacement: Impact on employment and livelihoods
  • Autonomous weapons: Military applications raise serious concerns
  • Sentience: If AI becomes sentient, what rights should it have?
  • Control: Ensuring humans remain in control of AI systems
  • Need for clear regulations and guidelines
  • Balancing innovation with responsible development

11. Exam Success Tips (Part 1)

💡 AI Types - Key Reminders
💡 AI Characteristics - Remember "LDA"
💡 Expert Systems - Structure
🧠 Memory Trick: "ERI" for Expert Systems
💡 Machine Learning vs Expert Systems

11. Exam Success Tips (Part 2)

❌ Common Mistakes to Avoid
💡 Impact Questions - Use This Structure
💡 Answer Structure Tips
🌟 Quick Reference Table
Topic Key Point
Weak AI Specific tasks only - what we have today
Strong AI Any human task - theoretical only
Expert System Knowledge base + Inference engine + User interface
Machine Learning Improves with experience/data
Neural Networks Pattern detection, fraud identification
Autonomous Robot Sensors + Processing + Actuators
📌 Final Exam Reminders