GARP RAI - Risk and Artificial Intelligence Exam
Page: 1 / 30
Total 150 questions
Question #1 (Topic: Exam A)
Modern AI capabilities have accelerated rapidly, driven by a combination of technological and methodological advances.
Which of the following is a key breakthrough contributing to the improvement in modern AI and machine learning performance?
Which of the following is a key breakthrough contributing to the improvement in modern AI and machine learning performance?
A. The introduction of heuristic search which reduced computation time for small optimization problems.
B. The shift from symbolic rule-based systems to expert systems that eliminated the need for large datasets.
C. The development of new training methods and availability of massive datasets enabled deep learning to scale.
D. The adoption of econometric models to improve interpretability of deep neural networks.
Answer: C
Question #2 (Topic: Exam A)
A corporate compliance team deployed a rule-based system to detect employee violations. The system worked well initially, but when employees used sophisticated techniques to avoid detection of violations, it consistently failed to flag them.
What fundamental limitation of classical AI does this reflect?
What fundamental limitation of classical AI does this reflect?
A. Its tendency to randomize outputs when encountering previously unseen inputs.
B. Its requirement for vast amounts of labeled training data to function properly.
C. Its inability to learn new patterns from data, restricting it to pre-defined knowledge.
D. Its reliance on deep neural networks that generate opaque reasoning paths.
Answer: C
Question #3 (Topic: Exam A)
Which of the following is considered a key breakthrough in Artificial Intelligence during the late 1980s?
A. The development of search, recursion, and look ahead algorithms.
B. The discovery and success of back-propagation.
C. The use of decision tree models for classification and prediction.
D. The development of neural network models with a single layer of artificial neurons.
Answer: B
Question #4 (Topic: Exam A)
AI systems like ChatGPT represent a major advancement in AI, and some feel these systems exhibit apparent artificial general intelligence.
Which of the following statements best describes how ChatGPT works?
Which of the following statements best describes how ChatGPT works?
A. ChatGPT responds to prompts by generating a response using a pre-determined menu of responses.
B. ChatGPT responds to prompts by imitating a response, with some variation, based on textual data it is trained on.
C. ChatGPT responds to prompts by generating a purely random response based on textual data it is trained on.
D. ChatGPT responds to prompts by generating a response that paraphrases the question it is asked.
Answer: B
Question #5 (Topic: Exam A)
Which of the statements below best describes how neural networks differ from classical AI?
A. Neural networks are modeled after neurons in brains, while classical AI is rule based.
B. Neural networks are constructed without any underlying structure, while classical AI is based on simple rules.
C. Neural networks are based on Markov chains, while classical AI is rule based.
D. Neural networks are modeled after electrical circuits, while classical AI is rule based.
Answer: A
