Artificial Intelligence MCQ

Q: Critique-then-improve in one prompt:

Correct answer: Makes the model list weaknesses, then rewrite fixing them. Structured self-editing raises quality without a second round-trip.

Q: Retrieval quality matters because:

Correct answer: Grounded answers can only be as good as the passages provided. RAG caps at retrieval quality — irrelevant passages produce grounded nonsense.

Q: The most durable way to understand the AI tool market is by:

Correct answer: Understanding tool categories and their uses. Categories persist while individual products come and go rapidly.

Q: The correct benchmark for evaluating a new AI tool is:

Correct answer: Your own real, representative tasks. Only your actual tasks reveal whether the tool improves your work.

Q: Before adopting an AI tool at work, the key privacy question is:

Correct answer: Where does my data go and is it used for training?. Data handling — storage, training use, business guarantees — is the core adoption risk.

AI concepts, machine learning, generative AI and prompt engineering.

Q71. Critique-then-improve in one prompt: Medium

Permalink →

Q72. Retrieval quality matters because: Medium

Permalink →

Q73. The most durable way to understand the AI tool market is by: Medium

Permalink →

Q74. The correct benchmark for evaluating a new AI tool is: Medium

Permalink →

Q75. Before adopting an AI tool at work, the key privacy question is: Medium

Permalink →

Q76. An AI research assistant differs from a plain chatbot mainly because it: Medium

Permalink →

Q77. When AI summarises a critical contract, you should: Medium

Permalink →

Q78. Recording meetings with an AI transcriber requires: Medium

Permalink →

Q79. Before using AI-generated images commercially you must check: Medium

Permalink →

Q80. The safe rule for AI-generated code is: Medium

Permalink →

🎯 Want a real test?

Take a timed quiz on this topic and see your rank on the leaderboard.

Browse Quizzes