Demystifying AI: AI, Machine Learning, Deep Learning & Generative AI | 100+ MCQs
Q1. What best defines Artificial Intelligence (AI)?
A. Automation of manual tasks
B. Simulation of intelligent behavior in machines
C. Storage of large datasets
D. Rule-based programming only
✅ Answer: B. Simulation of intelligent behavior in machines
Q2. Which capability distinguishes AI from traditional software systems?
A. Faster hardware
B. Ability to learn and adapt
C. Fixed rule execution
D. Deterministic outputs
✅ Answer: B. Ability to learn and adapt
Q3. AI occurs in which of the following?
A. Only machines
B. Only humans
C. Humans, animals, and some machines
D. Only robots
✅ Answer: C. Humans, animals, and some machines
Q4. Which of the following is NOT a core capability of AI?
A. Planning and decision-making
B. Knowledge representation
C. Emotional consciousness
D. Learning and adaptation
✅ Answer: C. Emotional consciousness
Q5. AI’s ability to interact with the real world primarily involves:
A. Data storage
B. Perception, understanding, and action
C. Compilation of code
D. Hardware acceleration
✅ Answer: B. Perception, understanding, and action
Q6. Which trait is unique to human intelligence compared to AI?
A. Fast computation
B. Large memory
C. Common sense and intuition
D. Pattern recognition
✅ Answer: C. Common sense and intuition
Q7. A limitation of human intelligence is:
A. Lack of creativity
B. Inability to reason
C. Limited memory capacity
D. Lack of judgment
✅ Answer: C. Limited memory capacity
Q8. Which is a strength of AI over humans?
A. Moral reasoning
B. Emotional intelligence
C. Fast processing of large data
D. Belief formation
✅ Answer: C. Fast processing of large data
Q9. AI systems lack which human capability inherently?
A. Pattern recognition
B. Learning
C. Common sense reasoning
D. Data processing
✅ Answer: C. Common sense reasoning
Q10. Which concern arises due to AI simulating human cognition?
A. Hardware scalability
B. Legal and ethical issues
C. Network latency
D. Power consumption
✅ Answer: B. Legal and ethical issues
Q11. Critical thinking primarily involves:
A. Acting on intuition
B. Evaluating evidence and arguments
C. Memorizing facts
D. Speed of decision-making
✅ Answer: B. Evaluating evidence and arguments
Q12. Identifying assumptions and biases is part of:
A. Emotional intelligence
B. Critical thinking
C. Automation
D. Reinforcement learning
✅ Answer: B. Critical thinking
Q13. Which behavior best reflects critical thinking?
A. Blindly trusting AI outputs
B. Revising beliefs based on evidence
C. Relying on gut feeling
D. Avoiding alternative views
✅ Answer: B. Revising beliefs based on evidence
Q14. Plausible reasoning differs from strict logic because it:
A. Avoids uncertainty
B. Handles incomplete information
C. Uses only mathematical proofs
D. Rejects assumptions
✅ Answer: B. Handles incomplete information
Q15. Why is critical thinking essential when using GenAI?
A. AI is deterministic
B. AI outputs are always correct
C. AI may hallucinate or amplify bias
D. AI replaces human judgment
✅ Answer: C. AI may hallucinate or amplify bias
Q16. Systems that think like humans focus on:
A. Rational optimization
B. Cognitive processes
C. Rule execution
D. Automation
✅ Answer: B. Cognitive processes
Q17. Systems that act rationally emphasize:
A. Mimicking emotions
B. Human-like appearance
C. Optimal goal achievement
D. Consciousness
✅ Answer: C. Optimal goal achievement
Q18. Chatbots that simulate conversation fall under:
A. Systems that think rationally
B. Systems that act like humans
C. Autonomous systems only
D. Superintelligence
✅ Answer: B. Systems that act like humans
Q19. Which AI view prioritizes logical correctness?
A. Think like humans
B. Act like humans
C. Think rationally
D. Act emotionally
✅ Answer: C. Think rationally
Q20. Self-driving cars are best classified as:
A. Thinking like humans
B. Acting rationally
C. Acting emotionally
D. Thinking irrationally
✅ Answer: B. Acting rationally
Q21. Siri or Alexa are examples of:
A. General AI
B. Super AI
C. Narrow AI
D. Conscious AI
✅ Answer: C. Narrow AI
Q22. General Intelligence AI refers to systems that:
A. Perform one task well
B. Think and decide like humans
C. Only automate workflows
D. Lack learning ability
✅ Answer: B. Think and decide like humans
Q23. Super Intelligence AI is best described as:
A. Currently deployed AI
B. AI equal to humans
C. Hypothetical AI surpassing humans
D. Rule-based systems
✅ Answer: C. Hypothetical AI surpassing humans
Q24. The concept of “singularity” relates to:
A. Narrow AI
B. Weak AI
C. Super Intelligence
D. Deterministic AI
✅ Answer: C. Super Intelligence
Q25. Narrow AI systems primarily:
A. Self-learn autonomously
B. Perform predefined tasks
C. Possess consciousness
D. Exhibit creativity
✅ Answer: B. Perform predefined tasks
Q26. Machine learning differs from traditional programming because it:
A. Uses more hardware
B. Learns from data instead of rules
C. Avoids algorithms
D. Eliminates computation
✅ Answer: B. Learns from data instead of rules
Q27. In traditional programming, output is produced by:
A. Data and learned models
B. Program and data
C. Random sampling
D. Neural networks
✅ Answer: B. Program and data
Q28. Supervised learning requires:
A. Unlabeled data
B. Reward signals only
C. Labeled data
D. No data
✅ Answer: C. Labeled data
Q29. Disease diagnosis using patient records is an example of:
A. Reinforcement learning
B. Unsupervised learning
C. Supervised learning
D. Random learning
✅ Answer: C. Supervised learning
Q30. Clustering is associated with:
A. Supervised learning
B. Reinforcement learning
C. Unsupervised learning
D. Deterministic learning
✅ Answer: C. Unsupervised learning
Q31. Deep learning is a subset of:
A. Statistics
B. Robotics
C. Machine Learning
D. Automation
✅ Answer: C. Machine Learning
Q32. Deep learning models are inspired by:
A. Databases
B. Human brain structure
C. Sorting algorithms
D. Expert systems
✅ Answer: B. Human brain structure
Q33. Deep learning automatically learns:
A. Handcrafted rules
B. High-level features
C. Program logic
D. Business policies
✅ Answer: B. High-level features
Q34. DL requires large datasets primarily because:
A. It is inefficient
B. It has many parameters
C. It avoids supervision
D. It ignores features
✅ Answer: B. It has many parameters
Q35. Since 2010, deep learning has outperformed ML especially in:
A. Accounting
B. Vision and speech
C. Database management
D. ERP systems
✅ Answer: B. Vision and speech
Q36. Generative AI primarily focuses on:
A. Classification
B. Prediction
C. Content creation
D. Optimization
✅ Answer: C. Content creation
Q37. ChatGPT is an example of:
A. Traditional AI
B. Rule-based AI
C. Generative AI
D. Expert systems
✅ Answer: C. Generative AI
Q38. GPT stands for:
A. General Processing Tool
B. Generative Pre-trained Transformer
C. Guided Prediction Technique
D. Global Programming Transformer
✅ Answer: B. Generative Pre-trained Transformer
Q39. ChatGPT gained rapid adoption mainly due to:
A. High hardware requirements
B. Intuitive interface and usefulness
C. Deterministic outputs
D. Low training data
✅ Answer: B. Intuitive interface and usefulness
Q40. LLMs primarily work by:
A. Understanding meaning like humans
B. Predicting the next token
C. Following fixed rules
D. Manual annotation
✅ Answer: B. Predicting the next token
Q41. A token in LLMs refers to:
A. A database record
B. A unit of text
C. A neuron
D. A rule
✅ Answer: B. A unit of text
Q42. LLMs are mainly trained using:
A. Supervised learning only
B. Reinforcement learning only
C. Unsupervised learning
D. Rule-based programming
✅ Answer: C. Unsupervised learning
Q43. RLHF is used to:
A. Increase model size
B. Reduce training data
C. Align outputs with human values
D. Eliminate bias completely
✅ Answer: C. Align outputs with human values
Q44. “Hallucination” in GenAI refers to:
A. Visual output
B. Accurate prediction
C. Fabricated information
D. Deterministic error
✅ Answer: C. Fabricated information
Q45. Larger parameter counts generally lead to:
A. Lower performance
B. Reduced capability
C. Improved model capacity
D. Deterministic outputs
✅ Answer: C. Improved model capacity
Q46. A major risk of GenAI is:
A. High speed
B. Bias amplification
C. Data compression
D. Deterministic logic
✅ Answer: B. Bias amplification
Q47. AI governance focuses on:
A. Hardware upgrades
B. Ethical and responsible use
C. Faster models
D. Removing humans
✅ Answer: B. Ethical and responsible use
Q48. Deterministic systems differ from probabilistic systems because they:
A. Use randomness
B. Provide fixed outputs
C. Learn continuously
D. Generate content
✅ Answer: B. Provide fixed outputs
Q49. GenAI is considered probabilistic because it:
A. Follows rules
B. Produces the same output
C. Generates multiple plausible outcomes
D. Avoids uncertainty
✅ Answer: C. Generates multiple plausible outcomes
Q50. The future role of AI in organizations is to:
A. Fully replace humans
B. Support and augment human work
C. Eliminate critical thinking
D. Remove governance
✅ Answer: B. Support and augment human work
Q51. Generative AI differs from traditional AI primarily because it:
A. Uses deterministic rules
B. Focuses on classification tasks
C. Generates novel content
D. Avoids neural networks
✅ Answer: C. Generates novel content
Q52. Which architecture is commonly used in modern Generative AI models?
A. Decision Trees
B. Support Vector Machines
C. Transformers
D. K-Means Clustering
✅ Answer: C. Transformers
Q53. GANs consist of which two components?
A. Encoder and Decoder
B. Generator and Discriminator
C. Predictor and Validator
D. Trainer and Tester
✅ Answer: B. Generator and Discriminator
Q54. Variational Autoencoders (VAEs) are primarily used for:
A. Classification
B. Compression only
C. Probabilistic data generation
D. Rule-based inference
✅ Answer: C. Probabilistic data generation
Q55. Which GenAI capability enables image creation from text prompts?
A. Speech Recognition
B. Reinforcement Learning
C. Multimodal learning
D. Clustering
✅ Answer: C. Multimodal learning
Q56. ChatGPT responses are generated based on:
A. Fixed templates
B. Statistical token prediction
C. Rule-based logic
D. Manual curation
✅ Answer: B. Statistical token prediction
Q57. Increasing the context window of an LLM enables:
A. Faster training
B. Longer input handling
C. Fewer parameters
D. Deterministic outputs
✅ Answer: B. Longer input handling
Q58. The randomness in LLM outputs is introduced to:
A. Reduce accuracy
B. Increase hardware usage
C. Enable diverse responses
D. Eliminate bias
✅ Answer: C. Enable diverse responses
Q59. Which factor most influences next-token prediction?
A. Hardware speed
B. Training corpus patterns
C. User location
D. Internet connectivity
✅ Answer: B. Training corpus patterns
Q60. LLMs appear to “understand” meaning because they:
A. Possess consciousness
B. Use symbolic reasoning
C. Learn statistical patterns
D. Follow ethical rules
✅ Answer: C. Learn statistical patterns
Q61. Parameters in an LLM represent:
A. Training datasets
B. Output tokens
C. Adjustable model weights
D. Ethical constraints
✅ Answer: C. Adjustable model weights
Q62. GPT-3 contains approximately:
A. 10 million parameters
B. 1 billion parameters
C. 175 billion parameters
D. 1 trillion parameters
✅ Answer: C. 175 billion parameters
Q63. The Pile dataset is best described as:
A. A labeled medical dataset
B. A curated finance dataset
C. A massive multi-domain corpus
D. A synthetic dataset
✅ Answer: C. A massive multi-domain corpus
Q64. RLHF primarily helps in:
A. Increasing model size
B. Improving alignment and safety
C. Speeding up inference
D. Eliminating hallucinations completely
✅ Answer: B. Improving alignment and safety
Q65. Constitutional AI aims to:
A. Replace governance
B. Encode ethical principles
C. Reduce data size
D. Improve hardware efficiency
✅ Answer: B. Encode ethical principles
Q66. Traditional AI is often referred to as:
A. Creative AI
B. Discriminative AI
C. Autonomous AI
D. Emotional AI
✅ Answer: B. Discriminative AI
Q67. Which task best suits traditional AI?
A. Writing poetry
B. Image synthesis
C. Fraud classification
D. Music composition
✅ Answer: C. Fraud classification
Q68. Generative AI outputs are best described as:
A. Binary
B. Deterministic
C. Synthetic and novel
D. Rule-bound
✅ Answer: C. Synthetic and novel
Q69. Compared to traditional AI, GenAI typically requires:
A. Less data
B. No training
C. Large unstructured data
D. Only labeled data
✅ Answer: C. Large unstructured data
Q70. A key advantage of GenAI is its ability to:
A. Replace humans
B. Scale creativity
C. Eliminate risk
D. Ensure accuracy
✅ Answer: B. Scale creativity
Q71. Using GenAI for fraud pattern simulation involves:
A. Classification
B. Data augmentation
C. Deterministic modeling
D. Manual rules
✅ Answer: B. Data augmentation
Q72. Sentiment analysis in finance is primarily used for:
A. Network security
B. Market forecasting
C. Hardware optimization
D. KYC verification
✅ Answer: B. Market forecasting
Q73. Personalized financial advice systems rely heavily on:
A. Batch processing
B. Reinforcement learning only
C. Customer data modeling
D. Manual intervention
✅ Answer: C. Customer data modeling
Q74. Which finance task benefits most from LLM-based automation?
A. ATM cash loading
B. Credit reports generation
C. Physical audits
D. Branch security
✅ Answer: B. Credit reports generation
Q75. GenAI in finance increases risk primarily due to:
A. Speed
B. Explainability challenges
C. Lower costs
D. Automation
✅ Answer: B. Explainability challenges
Q76. AI-generated ad copy demonstrates:
A. Deterministic logic
B. Creative synthesis
C. Manual segmentation
D. Static rules
✅ Answer: B. Creative synthesis
Q77. Customer journey mapping using GenAI enables:
A. Random targeting
B. Personalization at scale
C. Reduced engagement
D. Manual analytics
✅ Answer: B. Personalization at scale
Q78. Influencer marketing powered by GenAI primarily focuses on:
A. Cost reduction
B. Content authenticity simulation
C. Hardware efficiency
D. Data deletion
✅ Answer: B. Content authenticity simulation
Q79. AI-generated art challenges which traditional concept?
A. File storage
B. Intellectual property
C. Network bandwidth
D. Database indexing
✅ Answer: B. Intellectual property
Q80. Generative design in products enables:
A. Fixed templates
B. Exploration of multiple design options
C. Manual prototyping
D. Linear optimization
✅ Answer: B. Exploration of multiple design options
Q81. GenAI-assisted drug discovery primarily accelerates:
A. Manufacturing
B. Pattern discovery
C. Distribution
D. Regulation
✅ Answer: B. Pattern discovery
Q82. Medical imaging benefits from deep learning due to:
A. Low data needs
B. Feature hierarchy learning
C. Rule-based diagnosis
D. Determinism
✅ Answer: B. Feature hierarchy learning
Q83. Personalized learning systems use AI to:
A. Standardize content
B. Optimize learner paths
C. Eliminate teachers
D. Reduce assessment
✅ Answer: B. Optimize learner paths
Q84. Automated grading systems risk reducing:
A. Speed
B. Scalability
C. Human judgment
D. Consistency
✅ Answer: C. Human judgment
Q85. Cognitive offloading refers to:
A. Memory loss
B. Delegating routine tasks to AI
C. Multitasking
D. Over-reliance
✅ Answer: B. Delegating routine tasks to AI
Q86. Bias in AI systems often originates from:
A. Hardware design
B. Training data
C. Deployment speed
D. User interfaces
✅ Answer: B. Training data
Q87. AI hallucinations occur because models:
A. Are malicious
B. Optimize probabilities, not truth
C. Are deterministic
D. Use symbolic logic
✅ Answer: B. Optimize probabilities, not truth
Q88. Why must users apply critical thinking with AI outputs?
A. AI is slow
B. AI lacks context and judgment
C. AI avoids bias
D. AI is always correct
✅ Answer: B. AI lacks context and judgment
Q89. Which risk is most difficult to detect?
A. Hardware failure
B. Systemic bias
C. Latency
D. Storage limits
✅ Answer: B. Systemic bias
Q90. Data privacy risks increase when AI systems:
A. Run locally
B. Use open-source models
C. Process sensitive cloud data
D. Are deterministic
✅ Answer: C. Process sensitive cloud data
Q91. AI governance ensures:
A. Faster development
B. Responsible and ethical AI use
C. Elimination of humans
D. Unlimited automation
✅ Answer: B. Responsible and ethical AI use
Q92. Responsibility for AI risk management lies with:
A. Governments only
B. Developers only
C. Multi-stakeholder ecosystem
D. End users only
✅ Answer: C. Multi-stakeholder ecosystem
Q93. AI policies should primarily address:
A. Speed
B. Ethics, privacy, and accountability
C. Storage
D. Programming languages
✅ Answer: B. Ethics, privacy, and accountability
Q94. Continuous monitoring of AI systems is needed because:
A. Models never change
B. Risks evolve over time
C. Data is static
D. Outputs are deterministic
✅ Answer: B. Risks evolve over time
Q95. Oversight mechanisms help prevent:
A. Automation
B. Misuse and unintended harm
C. Innovation
D. Personalization
✅ Answer: B. Misuse and unintended harm
Q96. SLMs are best suited for:
A. Broad general intelligence
B. Domain-specific tasks
C. Large multilingual tasks
D. Public chatbots
✅ Answer: B. Domain-specific tasks
Q97. A key advantage of SLMs is:
A. Larger size
B. Higher latency
C. Lower cost and faster response
D. More hallucinations
✅ Answer: C. Lower cost and faster response
Q98. Local LLMs are preferred when:
A. Data sensitivity is high
B. Cloud access is unlimited
C. Latency is irrelevant
D. Scale is infinite
✅ Answer: A. Data sensitivity is high
Q99. Running LLMs locally improves:
A. Bias
B. Data privacy
C. Hallucinations
D. Randomness
✅ Answer: B. Data privacy
Q100. Open-source LLMs enable organizations to:
A. Eliminate governance
B. Customize models
C. Avoid responsibility
D. Reduce transparency
✅ Answer: B. Customize models
Q101. A foundation model is:
A. Task-specific only
B. A base model adaptable to many tasks
C. Deterministic system
D. Rule-based AI
✅ Answer: B. A base model adaptable to many tasks
Q102. LLMs are a subset of:
A. Narrow AI
B. Expert systems
C. Foundation models
D. Symbolic AI
✅ Answer: C. Foundation models
Q103. Probabilistic AI systems:
A. Always give the same output
B. Incorporate randomness
C. Avoid uncertainty
D. Use fixed rules
✅ Answer: B. Incorporate randomness
Q104. Deterministic systems are preferred when:
A. Creativity is needed
B. Predictability is critical
C. Language generation is required
D. Data is unstructured
✅ Answer: B. Predictability is critical
Q105. GenAI introduces which major shift in computing?
A. Reduced data
B. Non-deterministic behavior
C. Slower systems
D. Rule-based logic
✅ Answer: B. Non-deterministic behavior
Q106. AI should primarily be viewed as:
A. A replacement for humans
B. A support tool for human work
C. A standalone decision-maker
D. A risk-free system
✅ Answer: B. A support tool for human work
Q107. Successful GenAI adoption begins with:
A. Buying tools
B. Identifying high-value use cases
C. Hiring data scientists only
D. Eliminating manual work
✅ Answer: B. Identifying high-value use cases
Q108. Low-fidelity prototypes help organizations:
A. Finalize architecture
B. Visualize user experience early
C. Eliminate risk
D. Avoid feedback
✅ Answer: B. Visualize user experience early
Q109. User-centric design in GenAI ensures:
A. Technical elegance
B. Business value creation
C. Maximum automation
D. No governance
✅ Answer: B. Business value creation
Q110. AI strategy must align with:
A. Vendor roadmaps
B. Business goals and ethics
C. Hardware availability
D. Cloud pricing
✅ Answer: B. Business goals and ethics
Q111. Over-reliance on GenAI can weaken:
A. Hardware
B. Empathy and critical thinking
C. Data storage
D. Speed
✅ Answer: B. Empathy and critical thinking
Q112. Multitasking negatively affects:
A. Creativity only
B. Accuracy and memory
C. Automation
D. AI outputs
✅ Answer: B. Accuracy and memory
Q113. GenAI is most beneficial when it:
A. Replaces thinking
B. Automates routine tasks
C. Eliminates learning
D. Removes reflection
✅ Answer: B. Automates routine tasks
Q114. Tacit knowledge is best described as:
A. Explicit documentation
B. Subconscious human understanding
C. Database records
D. Model parameters
✅ Answer: B. Subconscious human understanding
Q115. AI cannot fully replace humans because it lacks:
A. Speed
B. Data
C. Contextual judgment
D. Computation
✅ Answer: C. Contextual judgment
Q116. The future of AI is best described as:
A. Fixed
B. Static
C. In flux
D. Fully predictable
✅ Answer: C. In flux
Q117. Business leaders must continuously:
A. Avoid AI
B. Learn and adapt to AI changes
C. Eliminate governance
D. Trust outputs blindly
✅ Answer: B. Learn and adapt to AI changes
Q118. AI delivers maximum value when combined with:
A. Automation alone
B. Human judgment
C. Deterministic rules
D. Minimal data
✅ Answer: B. Human judgment
Q119. Responsible AI use emphasizes:
A. Speed
B. Scale
C. Accountability
D. Replacement
✅ Answer: C. Accountability
Q120. The ultimate goal of AI in organizations is to:
A. Replace all jobs
B. Eliminate uncertainty
C. Augment human capability
D. Remove ethics
✅ Answer: C. Augment human capability
Q121. Which AI system is best for language-specific tasks with low latency?
A. Global LLM
B. SLM
C. Super AI
D. Rule engine
✅ Answer: B. SLM
Q122. AI systems without explainability are described as:
A. Transparent
B. Interpretable
C. Opaque
D. Manual
✅ Answer: C. Opaque
Q123. Autonomous systems differ from automated systems because they:
A. Support humans
B. Take over tasks completely
C. Require supervision
D. Follow scripts
✅ Answer: B. Take over tasks completely
Q124. Intelligent AI differs from instruments because it:
A. Follows rules
B. Mimics human cognition
C. Is deterministic
D. Avoids learning
✅ Answer: B. Mimics human cognition
Q125. The biggest misconception about LLMs is that they:
A. Use tokens
B. Learn from data
C. Truly understand meaning
D. Predict words
✅ Answer: C. Truly understand meaning
Q126. Ethical AI requires balancing innovation with:
A. Speed
B. Profit
C. Responsibility
D. Automation
✅ Answer: C. Responsibility
Q127. Bias amplification becomes dangerous when it is:
A. Visible
B. Systemic and hidden
C. Random
D. Manual
✅ Answer: B. Systemic and hidden
Q128. AI cannot replace leadership because leadership requires:
A. Data
B. Authority
C. Judgment and values
D. Algorithms
✅ Answer: C. Judgment and values
Q129. The most sustainable AI strategy is:
A. Tool-driven
B. Human-centered
C. Vendor-led
D. Speed-focused
✅ Answer: B. Human-centered
Q130. GenAI adoption without governance leads to:
A. Innovation
B. Trust
C. Risk exposure
D. Efficiency
✅ Answer: C. Risk exposure
Q131. Foundation models enable:
A. Single-task execution
B. Broad adaptability
C. Rule enforcement
D. Deterministic output
✅ Answer: B. Broad adaptability
Q132. Probabilistic AI systems are best suited for:
A. Mathematical proofs
B. Creative and language tasks
C. Sorting algorithms
D. Fixed workflows
✅ Answer: B. Creative and language tasks
Q133. Deterministic AI is ideal for:
A. Poetry generation
B. Medical dosage calculation
C. Image synthesis
D. Storytelling
✅ Answer: B. Medical dosage calculation
Q134. Which AI risk directly impacts trust?
A. Latency
B. Hallucinations
C. Storage
D. Compression
✅ Answer: B. Hallucinations
Q135. AI literacy is important because it:
A. Replaces governance
B. Enables informed usage
C. Reduces creativity
D. Eliminates bias
✅ Answer: B. Enables informed usage
Q136. Organizations should treat GenAI outputs as:
A. Absolute truth
B. Suggestions requiring validation
C. Final decisions
D. Deterministic results
✅ Answer: B. Suggestions requiring validation
Q137. The key differentiator between AI and humans remains:
A. Speed
B. Memory
C. Values and judgment
D. Data access
✅ Answer: C. Values and judgment
Q138. AI’s role in learning should emphasize:
A. Automation
B. Cognitive support
C. Replacement
D. Dependency
✅ Answer: B. Cognitive support
Q139. The success of AI depends most on:
A. Algorithms
B. Hardware
C. Human choices
D. Model size
✅ Answer: C. Human choices
Q140. AI strategy failure most often results from:
A. Poor UI
B. Lack of alignment with business goals
C. Excess data
D. Overtraining
✅ Answer: B. Lack of alignment with business goals
Q141. AI ethics primarily addresses:
A. Cost
B. Fairness, accountability, transparency
C. Speed
D. Architecture
✅ Answer: B. Fairness, accountability, transparency
Q142. Which capability remains uniquely human?
A. Pattern recognition
B. Contextual moral reasoning
C. Prediction
D. Optimization
✅ Answer: B. Contextual moral reasoning
Q143. AI hallucinations highlight the need for:
A. Faster models
B. Critical thinking
C. Larger datasets
D. Less randomness
✅ Answer: B. Critical thinking
Q144. AI adoption should be iterative because:
A. Models are static
B. Risks and capabilities evolve
C. Costs are fixed
D. Governance is optional
✅ Answer: B. Risks and capabilities evolve
Q145. Which principle best summarizes responsible GenAI use?
A. Replace humans
B. Trust outputs blindly
C. Augment, not replace
D. Eliminate judgment
✅ Answer: C. Augment, not replace
Q146. AI-supported decisions should always include:
A. Automation
B. Human oversight
C. Randomness
D. Speed
✅ Answer: B. Human oversight
Q147. The long-term value of AI lies in:
A. Model size
B. Human-AI collaboration
C. Full autonomy
D. Elimination of labor
✅ Answer: B. Human-AI collaboration
Q148. The biggest strategic risk of GenAI is:
A. Cost
B. Misaligned deployment
C. Speed
D. Adoption
✅ Answer: B. Misaligned deployment
Q149. AI literacy should be promoted across:
A. IT teams only
B. Leadership only
C. Entire organization
D. Vendors only
✅ Answer: C. Entire organization
Q150. The ultimate responsibility for AI outcomes rests with:
A. Algorithms
B. Vendors
C. Humans and institutions
D. Data
✅ Answer: C. Humans and institutions
Artificial Intelligence MCQs, Machine Learning MCQs, Deep Learning Questions, Generative AI MCQs, ChatGPT MBA Questions, LLM Concepts, AI Governance, AI Risks, Critical Thinking and AI, MBA AI Exam Questions
