Q1. Scenario: Describe three distinct real-world AI applications, one from healthcare, one from transportation, and one from entertainment. Explain the core AI technique used.
Healthcare: Medical imaging diagnosis (deep learning CNNs). Transportation: Autonomous driving (computer vision + reinforcement learning). Entertainment: Content recommendation (collaborative filtering, neural collaborative filtering).
Q2. Scenario: A fraud detection system in banking flags unusual transactions. How does it use machine learning? What is the trade-off between false positives and false negatives?
It trains on labeled legitimate/fraudulent transactions using anomaly detection or classification (e.g., random forest). High false positives inconvenience customers; false negatives cause financial loss. Adjust threshold accordingly.
Q3. Scenario: Explain how AI is used in agriculture (e.g., precision farming). Give two examples and the AI techniques involved.
1) Crop disease detection from drone images (computer vision, CNNs). 2) Soil moisture prediction for irrigation scheduling (time-series forecasting, LSTM).
Q4. Scenario: A social media platform uses AI to moderate content. What are the challenges (e.g., context understanding, free speech)? How might it fail?
Challenges: sarcasm, cultural nuances, evolving slang, and edge cases. Failure modes: false positives (censoring harmless posts), false negatives (missing hate speech), bias against certain groups.
Q5. Scenario: Voice assistants like Alexa are widely used. List three NLP tasks they perform, and explain one limitation of current technology.
Tasks: automatic speech recognition (ASR), natural language understanding (intent classification, entity extraction), text-to-speech (TTS). Limitation: lack of true conversational context across long dialogues.