UNIT 1: AI REFLECTION, PROJECT CYCLE & ETHICS
1. AI REFLECTION
What is AI Reflection?
AI Reflection means thinking about how Artificial Intelligence affects our lives, how machines behave, and how humans interact with AI systems.
Students reflect on:
- What AI can and cannot do
- How AI learns
- How AI impacts society
- How to use AI responsibly
Why is reflection important?
- Helps us understand AI’s abilities and limitations
- Improves decision making while using technology
- Encourages responsible and safe use of AI
- Helps in identifying ethical issues
Dimensions of AI (Three Domains)
1. Data Domain
AI that handles data: collecting, storing, and analyzing.
Examples:
- Google Maps using traffic data
- Amazon product recommendations
2. Computer Vision Domain
AI that can “see” and understand images or videos.
Examples:
- Face unlock in phones
- Self-driving cars detecting signals
3. Natural Language Processing (NLP) Domain
AI that understands and works with human language.
Examples:
- Chatbots
- Voice assistants (Alexa, Siri)
- Language translators
Types of AI (Simple classification)
1. Narrow AI (Weak AI)
- Designed for one specific task
- Most AI today belongs to this type
Examples: Google Search, YouTube recommendations
2. General AI
- AI that can think and perform any intellectual task like a human
- Does NOT exist yet
2. PROJECT CYCLE IN AI
The AI Project Cycle includes steps to solve a real-life problem using AI.
Step i: Problem Scoping
Understanding the problem clearly.
Includes:
- Identifying the problem
- Understanding why the problem exists
- Knowing who faces this problem
- Preparing problem statements
Example:
Traffic jam issue → Need an app that predicts traffic.
Step ii: Data Acquisition
Collecting data needed for solving the problem.
Sources:
- Surveys
- Sensors
- Websites
- Databases
Types of Data:
- Text
- Images
- Audio
- Video
Step iii: Data Exploration
Data Exploration
- Cleaning data
- Removing errors
- Understanding patterns
Step iv: Data Modelling
Modelling
- Training AI models using the collected data
- Choosing correct algorithms
- Testing the model’s performance
Step v: Evaluation
Checking:
- Does the model work correctly?
- Is the output accurate?
- Does it solve the problem?
If results are not good → improve data or model.
Step vi: Deployment
- Using the AI model in real life
- Integrating it into an app or website
- Monitoring its performance
3. AI ETHICS
What are ethics?
Ethics means rules and moral principles that guide right or wrong behaviour.
Why do we need ethics in AI?
- To ensure AI is safe and fair
- To protect people’s privacy
- To avoid misuse of technology
- To build trust between humans and machines
Key Ethical Principles
1. Transparency
AI systems should be clear about:
- How they work
- What data they use
- Why they give certain results
2. Fairness
- AI should treat everyone equally
- No discrimination based on gender, caste, colour, religion
3. Privacy
- Data must be protected
- Personal information should not be leaked or misused
4. Accountability
- Someone must take responsibility if an AI system causes harm
- Developers must ensure safe use
5. Safety & Security
- AI must not cause physical or digital harm
- Systems should be protected from hackers
Examples of Ethical Issues in AI
- A face-recognition system showing more errors for darker skin
- Social media apps collecting unnecessary personal data
- Biased AI recruitment systems rejecting qualified candidates
- Deepfake videos misused for harming people
Responsible Use of AI
Students should:
- Use AI tools for learning, not cheating
- Verify information before sharing
- Protect passwords and personal data
- Think before using or creating AI projects
SUMMARY
| Topic | Key Points |
|---|---|
| AI Reflection | Understanding AI’s impact, domains: Data, CV, NLP |
| Project Cycle | Problem Scoping → Data Acquisition →Data exploration→ Modelling → Evaluation → Deployment |
| AI Ethics | Transparency, Fairness, Privacy, Accountability, Safety |