Correct Answer: The ability to learn and solve problems
Explanation:
Intelligence includes learning, reasoning, problem-solving, and adapting to new situations.
Incorrect! Try again.
2Which activity is commonly associated with intelligence?
Concept of intelligence
Easy
A.Reasoning about a new problem
B.Turning off a device
C.Repeating the same movement
D.Displaying a fixed message
Correct Answer: Reasoning about a new problem
Explanation:
Reasoning about unfamiliar problems is an important part of intelligent behavior.
Incorrect! Try again.
3What does learning mean in the context of intelligence?
Concept of intelligence
Easy
A.Removing all stored information
B.Improving from experience
C.Avoiding every new situation
D.Following only one instruction
Correct Answer: Improving from experience
Explanation:
Learning allows an intelligent system to improve its performance using experience or information.
Incorrect! Try again.
4Which ability helps an intelligent system respond to changes in its environment?
Concept of intelligence
Easy
A.Duplication
B.Decoration
C.Compression
D.Adaptation
Correct Answer: Adaptation
Explanation:
Adaptation is the ability to adjust behavior when conditions or information change.
Incorrect! Try again.
5Which of the following is an example of human intelligence?
Concept of intelligence
Easy
A.Playing a fixed animation
B.Understanding a new language
C.Showing a constant screen image
D.Repeating a recorded sound
Correct Answer: Understanding a new language
Explanation:
Understanding a new language requires learning, interpretation, and reasoning.
Incorrect! Try again.
6What is weak AI designed to do?
Differences between weak AI and strong AI
Easy
A.Replace all human abilities
B.Think about every possible topic
C.Develop emotions like a person
D.Perform specific tasks
Correct Answer: Perform specific tasks
Explanation:
Weak AI, also called narrow AI, is created to perform particular tasks or solve specific problems.
Incorrect! Try again.
7Which is another name for weak AI?
Differences between weak AI and strong AI
Easy
A.Universal AI
B.Narrow AI
C.Human AI
D.General AI
Correct Answer: Narrow AI
Explanation:
Weak AI is commonly called narrow AI because it focuses on a limited task or area.
Incorrect! Try again.
8What is strong AI intended to achieve?
Differences between weak AI and strong AI
Easy
A.One automatic calculation
B.A single image filter
C.Human-like general intelligence
D.A fixed sorting procedure
Correct Answer: Human-like general intelligence
Explanation:
Strong AI refers to a theoretical system with general intelligence similar to human intelligence.
Incorrect! Try again.
9Which statement about today's common AI systems is correct?
Differences between weak AI and strong AI
Easy
A.Most understand every subject
B.Most are examples of weak AI
C.Most have human consciousness
D.Most can perform every human job
Correct Answer: Most are examples of weak AI
Explanation:
Most current AI applications are designed for specific tasks rather than general human-level intelligence.
Incorrect! Try again.
10A voice assistant that answers questions and sets reminders is usually classified as what type of AI?
Differences between weak AI and strong AI
Easy
A.Noncomputing intelligence
B.Weak AI
C.Strong AI
D.Biological intelligence
Correct Answer: Weak AI
Explanation:
A voice assistant performs a limited set of tasks, so it is an example of weak or narrow AI.
Incorrect! Try again.
11Who is often called the father of modern computer science and artificial intelligence?
History of AI from beginnings to present day
Easy
A.Isaac Newton
B.Alan Turing
C.Alexander Graham Bell
D.Charles Darwin
Correct Answer: Alan Turing
Explanation:
Alan Turing made foundational contributions to computing and ideas about machine intelligence.
Incorrect! Try again.
12What did the Turing Test propose evaluating?
History of AI from beginnings to present day
Easy
A.A robot's physical strength
B.A program's electrical voltage
C.A machine's ability to imitate human conversation
D.A computer's storage capacity
Correct Answer: A machine's ability to imitate human conversation
Explanation:
The Turing Test considers whether a machine can communicate in a way that is difficult to distinguish from a human.
Incorrect! Try again.
13At which 1956 event was the term artificial intelligence introduced?
History of AI from beginnings to present day
Easy
A.Boston Robotics Forum
B.Paris Science Fair
C.Dartmouth workshop
D.London Computing Expo
Correct Answer: Dartmouth workshop
Explanation:
The term artificial intelligence was introduced at the Dartmouth workshop in 1956.
Incorrect! Try again.
14What was an early goal of artificial intelligence research?
History of AI from beginnings to present day
Easy
A.Designing faster paper printers
B.Increasing battery color options
C.Building thinner television screens
D.Creating programs that solve problems
Correct Answer: Creating programs that solve problems
Explanation:
Early AI research focused on developing programs that could reason, solve problems, and perform tasks associated with human intelligence.
Incorrect! Try again.
15What is an AI winter?
History of AI from beginnings to present day
Easy
A.A period of reduced AI funding and interest
B.A holiday for studying intelligent machines
C.A method for cooling computer processors
D.A season for testing outdoor robots
Correct Answer: A period of reduced AI funding and interest
Explanation:
An AI winter is a period when progress expectations decline and investment in AI research decreases.
Incorrect! Try again.
16Which development has strongly supported recent progress in AI?
History of AI from beginnings to present day
Easy
A.Smaller research communities
B.Reduced internet access
C.More data and computing power
D.Fewer digital devices
Correct Answer: More data and computing power
Explanation:
Large datasets and powerful computing systems have helped modern AI models become more capable.
Incorrect! Try again.
17Which application commonly uses AI to identify objects in photographs?
Examples of today's real-world applications of AI
Easy
A.Mechanical drilling
B.Paper filing
C.Image recognition
D.Manual bookkeeping
Correct Answer: Image recognition
Explanation:
Image recognition systems use AI to identify objects, faces, or patterns in images.
Incorrect! Try again.
18Which AI application is commonly used by online stores?
Examples of today's real-world applications of AI
Easy
A.Handwritten invoices
B.Physical shelf painting
C.Product recommendations
D.Manual package wrapping
Correct Answer: Product recommendations
Explanation:
Recommendation systems analyze user behavior and suggest products that may be relevant.
Incorrect! Try again.
19What type of AI application can help banks identify unusual transactions?
Examples of today's real-world applications of AI
Easy
A.Weather forecasting
B.Photo editing
C.Fraud detection
D.Speech synthesis
Correct Answer: Fraud detection
Explanation:
Fraud detection systems look for unusual transaction patterns that may indicate suspicious activity.
Incorrect! Try again.
20Which technology uses AI to convert spoken words into written text?
Examples of today's real-world applications of AI
Easy
A.Speech recognition
B.Data encryption
C.Screen calibration
D.Image compression
Correct Answer: Speech recognition
Explanation:
Speech recognition systems process spoken language and produce text or commands.
Incorrect! Try again.
21A navigation system changes its route after detecting an unexpected road closure. Which aspect of intelligence is most clearly demonstrated?
Concept of intelligence
Medium
A.Following a fixed sequence of commands
B.Adapting behavior to new information
C.Storing previously observed routes
D.Displaying directions on a screen
Correct Answer: Adapting behavior to new information
Explanation:
The system uses new information about the road closure to modify its behavior, demonstrating adaptation and problem-solving.
Incorrect! Try again.
22Why is the ability to generalize important for an intelligent system?
Concept of intelligence
Medium
A.It limits the system to examples seen during its training
B.It enables the system to operate without receiving any data
C.It allows the system to apply learned patterns to new situations
D.It guarantees that the system will never make an incorrect decision
Correct Answer: It allows the system to apply learned patterns to new situations
Explanation:
Generalization allows an intelligent system to use knowledge from previous examples when encountering unfamiliar but related situations.
Incorrect! Try again.
23A program wins a chess game by evaluating possible moves but cannot explain everyday objects or understand spoken language. What does this example suggest about intelligence?
Concept of intelligence
Medium
A.Intelligence can be strong in one domain and limited in others
B.A system cannot be intelligent unless it understands language
C.Successful performance in one task proves human-level intelligence
D.Chess performance requires abilities from every area of intelligence
Correct Answer: Intelligence can be strong in one domain and limited in others
Explanation:
Intelligent behavior may be specialized. A system can perform well in a narrow task without possessing broad, general abilities.
Incorrect! Try again.
24Which situation best demonstrates reasoning rather than simple memorization?
Concept of intelligence
Medium
A.Repeating a stored definition of artificial intelligence
B.Selecting a safe route by comparing traffic, distance, and weather
C.Recognizing a photograph previously included in training data
D.Listing all commands supported by a software application
Correct Answer: Selecting a safe route by comparing traffic, distance, and weather
Explanation:
The system must combine several factors and evaluate alternatives, which requires reasoning rather than merely retrieving stored information.
Incorrect! Try again.
25A robot correctly identifies a cup in bright light but fails when the same cup is partly hidden. Which capability would most improve its intelligence in this situation?
Concept of intelligence
Medium
A.A larger collection of motor commands
B.More frequent repetition of the robot's output
C.Robust perception under changing conditions
D.Faster execution of the same visual rule
Correct Answer: Robust perception under changing conditions
Explanation:
The robot needs to recognize relevant objects despite changes such as occlusion and lighting.
Incorrect! Try again.
26A customer-service chatbot answers questions about product returns but cannot perform unrelated tasks. How should it be classified?
Differences between weak AI and strong AI
Medium
A.Strong AI because it communicates with customers
B.Weak AI because it is designed for a specific task
C.Neither type because chatbots cannot use artificial intelligence
D.Strong AI because it uses natural-language processing
Correct Answer: Weak AI because it is designed for a specific task
Explanation:
Weak AI, also called narrow AI, is built to perform particular tasks without possessing general human-like intelligence.
Incorrect! Try again.
27Which statement best describes the intended meaning of strong AI?
Differences between weak AI and strong AI
Medium
A.A system that demonstrates general intelligence comparable to human intelligence
B.A system that performs one task with very high accuracy
C.A system that uses many sensors to collect information
D.A system that produces outputs faster than traditional software
Correct Answer: A system that demonstrates general intelligence comparable to human intelligence
Explanation:
Strong AI refers to the theoretical development of systems with broad, human-like understanding and intelligence across many tasks.
Incorrect! Try again.
28A medical image system detects lung abnormalities accurately but cannot diagnose unrelated conditions or hold a general conversation. What conclusion is most appropriate?
Differences between weak AI and strong AI
Medium
A.It is weak AI despite its high performance in one area
B.It is strong AI because medical decisions require reasoning
C.It is not AI because its task is limited to image analysis
D.It is strong AI because its accuracy exceeds that of some humans
Correct Answer: It is weak AI despite its high performance in one area
Explanation:
High accuracy in a specialized task does not make a system generally intelligent; the system remains narrow or weak AI.
Incorrect! Try again.
29Which capability would provide the strongest evidence that a system is approaching strong AI rather than remaining weak AI?
Differences between weak AI and strong AI
Medium
A.Generating several answers to the same question
B.Storing a larger database of task-specific examples
C.Classifying images faster than previous systems
D.Learning and transferring knowledge across unrelated tasks
Correct Answer: Learning and transferring knowledge across unrelated tasks
Explanation:
General intelligence involves applying knowledge flexibly across different domains rather than improving only within one narrow task.
Incorrect! Try again.
30Why can a voice assistant appear intelligent without necessarily being strong AI?
Differences between weak AI and strong AI
Medium
A.It must possess consciousness whenever it produces natural sentences
B.It automatically understands every subject as well as a human expert
C.It may combine several specialized systems for speech and language tasks
D.It cannot use algorithms because speech is a human-only ability
Correct Answer: It may combine several specialized systems for speech and language tasks
Explanation:
A voice assistant can perform complex but limited functions through specialized components without having general understanding or consciousness.
Incorrect! Try again.
31Why is the 1950 Dartmouth workshop commonly associated with the beginning of AI as a formal field?
History of AI from beginnings to present day
Medium
A.It introduced the first commercially available personal computer
B.It proposed studying machine intelligence as a distinct research area
C.It created the first large-scale internet search engine
D.It demonstrated the first successful autonomous vehicle on public roads
Correct Answer: It proposed studying machine intelligence as a distinct research area
Explanation:
The Dartmouth workshop helped establish artificial intelligence as an organized academic field and popularized the term.
Incorrect! Try again.
32What was a major characteristic of early symbolic AI systems?
History of AI from beginnings to present day
Medium
A.They generated predictions using modern transformer architectures
B.They learned all their knowledge from massive labeled datasets
C.They relied mainly on explicitly coded rules and logical representations
D.They depended primarily on cloud-based graphical processing units
Correct Answer: They relied mainly on explicitly coded rules and logical representations
Explanation:
Early symbolic AI represented knowledge with rules, symbols, and logic rather than learning primarily from large datasets.
Incorrect! Try again.
33Expert systems became popular because they could provide useful decisions in limited domains. What was a common limitation of these systems?
History of AI from beginnings to present day
Medium
A.They were unable to produce explanations for their conclusions
B.They lacked any method for representing domain knowledge
C.They could only run on mobile devices with internet access
D.They struggled to handle situations not covered by their programmed rules
Correct Answer: They struggled to handle situations not covered by their programmed rules
Explanation:
Rule-based expert systems depended heavily on predefined knowledge, so they often performed poorly in unfamiliar or exceptional situations.
Incorrect! Try again.
34What factor most directly contributed to the recent growth of machine-learning applications?
History of AI from beginnings to present day
Medium
A.Improved access to data, computing power, and learning algorithms
B.Reduced availability of digital information
C.A complete replacement of mathematical models by human judgment
D.A shift away from automated decision-making in industry
Correct Answer: Improved access to data, computing power, and learning algorithms
Explanation:
Modern AI development has been accelerated by large datasets, powerful hardware, and improved machine-learning methods.
Incorrect! Try again.
35AI winters occurred when enthusiasm and funding declined. Which situation could most likely contribute to an AI winter?
History of AI from beginnings to present day
Medium
A.Research results fail to meet promises, reducing confidence and investment
B.Systems consistently exceed ambitious public expectations
C.New hardware makes large-scale experiments more affordable
D.Applications demonstrate reliable value across several industries
Correct Answer: Research results fail to meet promises, reducing confidence and investment
Explanation:
AI winters have been associated with unmet expectations, disappointing results, and reduced funding or public confidence.
Incorrect! Try again.
36A bank uses an AI model to flag unusual credit-card transactions for review. What is the primary AI application?
Examples of today's real-world applications of AI
Medium
A.Image compression
B.Route planning
C.Fraud detection
D.Speech synthesis
Correct Answer: Fraud detection
Explanation:
Identifying unusual transaction patterns to detect possible unauthorized activity is a common use of AI for fraud detection.
Incorrect! Try again.
37An online store recommends products based on a customer's previous purchases and browsing history. Which technique is being applied?
Examples of today's real-world applications of AI
Medium
A.Optical character recognition
B.Industrial automation
C.Recommender systems
D.Autonomous navigation
Correct Answer: Recommender systems
Explanation:
Recommender systems analyze user behavior and item characteristics to suggest products that may interest the user.
Incorrect! Try again.
38A hospital uses an AI tool to identify possible tumors in medical scans, but a radiologist makes the final decision. What is the main benefit of this arrangement?
Examples of today's real-world applications of AI
Medium
A.AI removes the need for medical training and clinical review
B.AI replaces all medical responsibility with automated predictions
C.AI guarantees that every scan will receive the same diagnosis
D.AI supports professional judgment while retaining human oversight
Correct Answer: AI supports professional judgment while retaining human oversight
Explanation:
AI can help identify patterns and prioritize cases, while qualified professionals interpret results and remain responsible for final decisions.
Incorrect! Try again.
39Which example best illustrates natural-language processing in a current application?
Examples of today's real-world applications of AI
Medium
A.A system adjusting a factory arm to a fixed position
B.A system measuring temperature with a digital sensor
C.A system translating a customer message from one language into another
D.A system increasing screen brightness in direct sunlight
Correct Answer: A system translating a customer message from one language into another
Explanation:
Language translation requires a system to analyze and generate human language, making it an application of natural-language processing.
Incorrect! Try again.
40A logistics company uses AI to predict delivery delays by analyzing weather, traffic, and historical shipment data. What makes this an AI application rather than a simple fixed schedule?
Examples of today's real-world applications of AI
Medium
A.It assigns every shipment the same expected arrival time
B.It displays delivery information using a digital interface
C.It follows a schedule that cannot change after publication
D.It uses changing data to estimate outcomes and support decisions
Correct Answer: It uses changing data to estimate outcomes and support decisions
Explanation:
The system analyzes multiple data sources to make predictions and adapt decisions, which is characteristic of AI-based forecasting.
Incorrect! Try again.
41An agent performs extremely well on a fixed set of chess positions but cannot recognize a chessboard when its orientation changes. Which conclusion is most defensible?
Concept of intelligence
Hard
A.It demonstrates human-level understanding
B.It demonstrates broad intelligence
C.It demonstrates task-specific competence
D.It lacks any form of intelligence
Correct Answer: It demonstrates task-specific competence
Explanation:
Performance on a narrowly defined task can indicate specialized competence without establishing flexible, general intelligence.
Incorrect! Try again.
42A home robot learns to navigate unfamiliar rooms, asks for clarification when instructions are ambiguous, and improves after observing failed attempts. Which combination most strongly supports attributing intelligence to the robot?
Concept of intelligence
Hard
A.Human-like appearance and emotional speech
B.Preprogrammed responses and fixed sensors
C.Adaptation, reasoning, and goal-directed behavior
D.Large memory and rapid computation
Correct Answer: Adaptation, reasoning, and goal-directed behavior
Explanation:
Flexible adaptation, inference under uncertainty, and purposeful action are stronger indicators of intelligence than appearance, speed, or storage capacity.
Incorrect! Try again.
43Why is passing a conversational imitation test insufficient by itself to establish that a system possesses human-like intelligence?
Concept of intelligence
Hard
A.A system may imitate linguistic behavior without general understanding
B.The test measures only physical coordination
C.Conversation never requires information processing
D.Human judges cannot evaluate language behavior
Correct Answer: A system may imitate linguistic behavior without general understanding
Explanation:
A system can produce convincing conversational outputs through statistical or scripted behavior while lacking transfer, grounding, or broad conceptual understanding.
Incorrect! Try again.
44Two agents achieve the same reward in a changing environment. Agent A follows a massive lookup table, while Agent B infers compact rules and succeeds in novel situations. Which assessment is strongest?
Concept of intelligence
Hard
A.Agent B shows stronger evidence of generalization
B.Neither agent can be considered goal-directed
C.Agent A is necessarily more intelligent
D.Both agents are equally intelligent in every sense
Correct Answer: Agent B shows stronger evidence of generalization
Explanation:
Success on previously unseen situations using inferred structure provides stronger evidence of flexible intelligence than memorized responses.
Incorrect! Try again.
45A medical model predicts disease accurately in one hospital but fails after a change in patient demographics and equipment. Which property of intelligence is most clearly missing?
Concept of intelligence
Hard
A.Data storage capacity
B.Environmental robustness
C.Arithmetic precision
D.Response speed
Correct Answer: Environmental robustness
Explanation:
Intelligent behavior should remain reasonably effective under relevant changes in context, rather than depending entirely on one narrow distribution.
Incorrect! Try again.
46Which statement best distinguishes weak AI from strong AI without relying on whether the system is useful or commercially successful?
Differences between weak AI and strong AI
Hard
A.Weak AI solves specified tasks, whereas strong AI implies general understanding
B.Weak AI uses software, whereas strong AI uses hardware
C.Weak AI is always inaccurate, whereas strong AI is always accurate
D.Weak AI is statistical, whereas strong AI is exclusively symbolic
Correct Answer: Weak AI solves specified tasks, whereas strong AI implies general understanding
Explanation:
Weak AI refers to task-oriented systems; strong AI is associated with general, human-like understanding or cognition rather than a particular implementation method.
Incorrect! Try again.
47A language model writes a persuasive legal argument but cannot reliably explain the factual consequences of a new law outside its training experience. What does this most directly illustrate?
Differences between weak AI and strong AI
Hard
A.Weak AI cannot generate novel language
B.Strong AI necessarily requires a physical body
C.Fluent performance does not prove general intelligence
D.Any accurate output proves conscious understanding
Correct Answer: Fluent performance does not prove general intelligence
Explanation:
High-quality output in one domain does not establish broad transfer, grounded understanding, or the general reasoning associated with strong AI.
Incorrect! Try again.
48A factory system detects defects, schedules maintenance, and adjusts production, but all capabilities are separately engineered for factory operations. How should it be classified most cautiously?
Differences between weak AI and strong AI
Hard
A.Weak AI because its competence remains domain-bounded
B.Strong AI because it adapts to sensor data
C.Neither AI because it operates without emotions
D.Strong AI because it performs several tasks
Correct Answer: Weak AI because its competence remains domain-bounded
Explanation:
Multiple specialized functions do not amount to strong AI when the system lacks general-purpose understanding beyond its designed operational domain.
Incorrect! Try again.
49Which observation would provide the strongest evidence for a claim that a system approaches strong AI, while still not proving consciousness?
Differences between weak AI and strong AI
Hard
A.It produces a human-like voice
B.It transfers concepts across unfamiliar domains and tasks
C.It classifies images faster than humans
D.It stores more facts than a person
Correct Answer: It transfers concepts across unfamiliar domains and tasks
Explanation:
Broad transfer and flexible problem-solving are more relevant to general intelligence than speed, voice, or factual storage; consciousness remains a separate issue.
Incorrect! Try again.
50Why can a collection of many weak AI systems still fail to constitute strong AI?
Differences between weak AI and strong AI
Hard
A.Weak systems cannot process numerical inputs
B.Weak systems are incapable of interacting with one another
C.Strong AI requires every module to be symbolic
D.Combining narrow modules does not guarantee unified general understanding
Correct Answer: Combining narrow modules does not guarantee unified general understanding
Explanation:
A system may coordinate many specialized capabilities yet lack common-sense reasoning, autonomous transfer, and integrated understanding across contexts.
Incorrect! Try again.
51Which historical sequence is chronologically correct?
History of AI from beginnings to present day
Hard
A.Expert systems, Dartmouth workshop, deep learning, perceptrons
B.Deep learning, perceptrons, Dartmouth workshop, expert systems
D.Dartmouth workshop, expert systems, AI winter, modern deep learning
Correct Answer: Dartmouth workshop, expert systems, AI winter, modern deep learning
Explanation:
The 1956 Dartmouth workshop preceded expert-system growth, later funding setbacks and AI winters, and the recent deep-learning expansion.
Incorrect! Try again.
52What was a major reason early symbolic AI systems struggled when deployed in broad, real-world environments?
History of AI from beginnings to present day
Hard
A.They lacked all forms of logical reasoning
B.They depended on explicit rules that handled exceptions poorly
C.They were unable to represent any domain knowledge
D.They required neural networks for every decision
Correct Answer: They depended on explicit rules that handled exceptions poorly
Explanation:
Symbolic systems could reason effectively in constrained domains, but the cost of encoding commonsense knowledge and exceptions grew rapidly.
Incorrect! Try again.
53The success of modern deep learning is best explained by which combination of historical developments?
History of AI from beginnings to present day
Hard
A.A complete rejection of statistics and optimization
B.Large datasets, increased computation, and improved algorithms
C.The disappearance of uncertainty from real-world tasks
D.Less data, weaker hardware, and fewer applications
Correct Answer: Large datasets, increased computation, and improved algorithms
Explanation:
Deep learning benefited from scalable computation, abundant digital data, improved architectures, and advances in optimization and training methods.
Incorrect! Try again.
54Why are AI winters important when interpreting claims about current AI progress?
History of AI from beginnings to present day
Hard
A.They demonstrate that hardware has no effect on AI
B.They reveal how expectations and funding can exceed practical capabilities
C.They show that AI research permanently ended
D.They prove symbolic methods were never useful
Correct Answer: They reveal how expectations and funding can exceed practical capabilities
Explanation:
AI winters followed periods in which ambitious promises were not matched by robust performance, emphasizing the danger of hype-driven evaluation.
Incorrect! Try again.
55Which development most directly enabled AI systems to learn representations from large collections of unstructured data rather than depend entirely on hand-crafted features?
History of AI from beginnings to present day
Hard
A.The spread of mechanical calculators
B.The replacement of algorithms by human judgment
C.Representation learning with deep neural networks
D.The elimination of training data requirements
Correct Answer: Representation learning with deep neural networks
Explanation:
Deep neural networks can learn hierarchical features from raw or minimally processed data, reducing reliance on manually designed representations.
Incorrect! Try again.
56A bank uses an AI model to approve loans. The model is highly accurate overall but rejects applicants from one region at twice the rate of similarly qualified applicants elsewhere. What is the primary issue?
Examples of today's real-world applications of AI
Hard
A.Low computational efficiency
B.Excessive model interpretability
C.Potential subgroup fairness and distributional bias
D.Insufficient database normalization
Correct Answer: Potential subgroup fairness and distributional bias
Explanation:
Large aggregate accuracy can conceal systematically unequal error rates or decisions across demographic or geographic groups.
Incorrect! Try again.
57An autonomous vehicle identifies a pedestrian correctly in clear weather but misses one in heavy rain. Which evaluation is most appropriate?
Examples of today's real-world applications of AI
Hard
A.Test performance across relevant environmental conditions
B.Measure only the vehicle's processing speed
C.Evaluate only its average accuracy in clear weather
D.Remove difficult weather cases from the deployment scope
Correct Answer: Test performance across relevant environmental conditions
Explanation:
Safety-critical AI must be evaluated under the range of conditions it may encounter, especially those that alter sensor quality or risk.
Incorrect! Try again.
58A hospital deploys an AI diagnostic tool that performs well in trials but gives unsafe recommendations when clinicians treat its confidence score as certainty. Which safeguard is most directly relevant?
Examples of today's real-world applications of AI
Hard
A.Train only on synthetic images
B.Remove all clinical oversight
C.Provide calibrated uncertainty and human review
D.Increase the display brightness
Correct Answer: Provide calibrated uncertainty and human review
Explanation:
Communicating uncertainty and preserving qualified oversight can reduce automation bias and prevent probabilistic predictions from being treated as guarantees.
Incorrect! Try again.
59A recommendation system increases user engagement by repeatedly showing sensational content, while reliable information receives less exposure. Which trade-off is being illustrated?
Examples of today's real-world applications of AI
Hard
A.Engagement optimization versus information quality
B.Latency versus memory allocation
C.Supervision versus sensor calibration
D.Encryption versus model compression
Correct Answer: Engagement optimization versus information quality
Explanation:
Optimizing a measurable proxy such as engagement can produce harmful content patterns when the proxy does not fully represent quality or social value.
Incorrect! Try again.
60An AI hiring system was trained on historical employee records from a workforce that was already demographically unbalanced. What is the most likely technical concern?
Examples of today's real-world applications of AI
Hard
A.Historical patterns may be reproduced as biased predictions
B.Training data automatically guarantees fairness
C.The model cannot process categorical variables
D.A larger model always removes selection bias
Correct Answer: Historical patterns may be reproduced as biased predictions
Explanation:
Models learn statistical regularities from their data; if historical outcomes reflect unequal opportunities, those patterns can be encoded and amplified.
Incorrect! Try again.
Did this save you a night before the exam?
LPU Notes is free, and it stays free. Ads cover part of the server bill.
The rest comes out of a student's own pocket: the domain, the storage,
and keeping the site up through the weeks everyone needs it at once.
The payment button didn't load. An ad blocker or a filtered network is the usual reason.
to try again.
Nothing here is ever locked, and nothing unlocks. Chip in only if it was worth it.
What it pays for →