Page 211 - Artificial Intellegence_v2.0_Class_11
P. 211
3. Awareness is a key factor that plays an important role in reducing the trust gap.
4. AI bias leads to a skewed output.
5. Humans and machines should not work together to reduce bias.
SECTION B (Subjective Type Questions)
A. Short answer type questions.
1. What is societal bias?
2. Give two instances of bias in AI systems.
3. How can we reduce bias in AI?
4. Mention two ethical issues concerning driverless cars.
5. Define AI bias.
B. Long answer type questions.
1. Explain briefly the four features that can help improve AI's public perception and reliability.
2. Mention few ethical concerns around AI.
3. Should AI be granted the full access to data collected from smartphones? Discuss.
4. How can you prevent algorithmic bias?
5. How can we build trust in AI?
C. Competency-based/Application-based questions:
1. Everyone is using ChatGPT nowadays. We all know that ChatGPT conversations are recorded and used to train new
models. Due to this, ChatGPT may duplicate any personal information—or fake information—that a user enters
about themselves or another person in later outputs. List 2 other ethical concerns that you feel ChatGPT may have.
2. “The existence of an AI that is completely unbiased is theoretically possible but unlikely.” Comment.
In Life Life Skills & Values
Discuss the possible bias we have/face in our daily lives, e.g., avoid haircut on a particular day of the week,
seeing black cat is a bad omen etc. Do you feel that such biases should no longer have a place in our lives?
Why/Why not?
Deep Thinking Experiential Learning
Read the article - Can a machine learn morality?
https://economictimes.indiatimes.com/news/science/can-a-machine-learn-morality/
articleshow/87818790.cms?from=mdr
Then visit Ask Delphi - https://delphi.allenai.org/
Discuss among students your own views on Delphi. The following questions may help you frame your views:
• Do you feel this AI has still a lot to learn? Why/Why not?
• If your answer to the above question is ‘Yes’, what kind of data should be added to Delphi’s AI model?
• Do you think such moral machines should be integrated in our justice system? Why/Why not?
AI Values (Ethical Decision Making) 209

