Unit 1: Foundations of Artificial Intelligence - Practice Quiz

1 What is the main goal of Artificial Intelligence?

2 Which event is commonly associated with the birth of AI as a formal field?

3 What type of AI is designed to perform one specific task?

4 What does Artificial General Intelligence refer to?

5 What is Machine Learning?

6 Deep Learning mainly uses which type of model?

7 Which field combines statistics, programming, and domain knowledge to study data?

8 What is the first step in many AI problem-solving processes?

9 What is the purpose of training an AI model?

10 Which stage checks how well an AI model performs on data?

11 What is an intelligent agent?

12 In an intelligent agent, what do sensors do?

13 What do actuators allow an agent to do?

14 How can AI support healthcare?

15 How is AI commonly used in manufacturing?

16 Which is an example of AI in a smart city?

17 What does fairness in AI aim to reduce?

18 Why is transparency important in AI systems?

19 What can Generative AI create?

20 What is a prompt in Generative AI?

21 A bank uses a computer system that examines previous loan decisions and recommends whether a new applicant should be approved. Which development in AI does this example best represent?

22 Why did early AI systems often perform poorly when faced with situations outside their programmed rules?

23 A customer-service chatbot can answer questions about billing but cannot control a robot or diagnose medical conditions. How should this system be classified?

24 Which situation would provide the strongest evidence that a system is approaching Artificial General Intelligence?

25 A team builds a model that predicts customer churn using customer records, visualizes trends, and recommends retention strategies. Which description best fits the overall work?

26 A neural network with many hidden layers learns features directly from thousands of medical images. Which term most specifically describes this approach?

27 Which statement best explains the relationship among AI, machine learning, and deep learning?

28 An AI team discovers that its training data contains many duplicate records and inconsistent labels. Which lifecycle stage should address this issue first?

29 A company wants an AI system to identify defective products. What should the team define before selecting an algorithm?

30 A model achieves high accuracy during training but performs poorly on new customer records. What is the most likely problem?

31 A navigation system must choose among several routes while considering travel time, traffic, and road closures. Which AI problem-solving activity is most relevant?

32 A vacuum-cleaning agent senses dirt, location, and obstacles, then chooses actions to clean rooms efficiently. What makes it an intelligent agent?

33 For an autonomous delivery drone operating in a city, which environmental property makes decision-making especially challenging?

34 A medical diagnosis agent recommends tests and treatments based on symptoms, patient history, and test results. Which component defines what the agent is trying to achieve?

35 A factory robot detects a worker nearby and slows its movement to reduce the risk of injury. Which AI application is demonstrated?

36 An AI healthcare system flags patients who may develop a disease, but a physician makes the final decision. What is the main advantage of this arrangement?

37 A manufacturing system predicts when a machine is likely to fail by analyzing vibration and temperature readings. What type of application is this?

38 A smart-city platform adjusts traffic signals using live vehicle-flow data. Which outcome is the system primarily designed to improve?

39 A hiring model selects candidates from one demographic group at a much higher rate because its training data reflected past biased decisions. Which responsible AI principle is most directly involved?

40 Which practice would best improve the accountability of an AI system used to approve insurance claims?

41 A system uses symbolic rules to diagnose equipment failures, but its rules are manually authored and never updated from operational data. Which limitation most directly distinguishes it from modern learning-based AI?

42 Why did expert systems experience a decline in practical adoption despite achieving strong performance in narrow domains?

43 A hospital system can diagnose diseases, schedule staff, and optimize treatment plans, but only within healthcare tasks for which it was designed. How should this system be classified?

44 Which observation would provide the strongest evidence that a system is not merely reactive but has a limited model-based capability?

45 A team cleans customer data, engineers variables, trains a gradient-boosted model, evaluates fairness, and communicates business recommendations. Which description best captures the overall work?

46 Which statement correctly distinguishes deep learning from machine learning?

47 A fraud model performs well on historical transactions but fails after criminals change their behavior. Which lifecycle activity is most important for detecting and addressing this issue?

48 A loan approval model is trained using a feature that encodes decisions made by previous loan officers. The model scores highly in validation. What is the primary concern?

49 For an autonomous delivery robot, the requirement is to minimize late deliveries while ensuring that collisions remain below a strict safety threshold. Which formulation is most appropriate?

50 A model achieves 98% accuracy when 98% of cases are negative, but it detects only 20% of positive cases. Which evaluation change is most justified?

51 A warehouse robot receives noisy camera observations, cannot observe every aisle, and must account for actions that affect future locations. Which environment properties apply?

52 Why is a rational agent not necessarily guaranteed to choose the action that produces the best eventual outcome?

53 A thermostat turns heating on whenever the current temperature is below a threshold, without storing previous readings or modeling room heat loss. What type of agent is it closest to?

54 A surgical-assistance robot detects a rare anatomical configuration that differs from its training data. Which design most appropriately reduces safety risk?

55 A predictive-maintenance system reports fewer machine failures after deployment, but operators also began replacing components more frequently. What must be analyzed before concluding that the model improved reliability?

56 A smart-city traffic model reduces average travel time but increases delays in neighborhoods with limited sensor coverage. Which intervention best addresses the identified deployment risk?

57 A hiring model has equal overall accuracy for two demographic groups, but its false-negative rate is substantially higher for one group. What does this demonstrate?

58 Which practice most directly improves auditability for a high-impact AI system after it has been deployed?

59 An AI provider claims that a model is fair because it does not use race as an input. Why is this claim insufficient?

60 A language model produces a confident but unsupported citation. Which explanation is most accurate?