• Practice
  • About 25 min
  • 15 practice questions

Final self-assessment: Generative AI and prompting

Test yourself on the whole course: 15 questions with answers explained.

If something looks wrong, please tell us.

Attempt all 15 questions in one sitting of about 25 minutes before looking at any answers. They mix ideas from all five lessons and are harder than the lesson checks, with several scenario questions. Mark your answers only after you finish, and read every explanation, including for the ones you got right.

Final self-assessment

  1. 1. Which statement about large language models is correct?

    1. They store a table of verified facts and quote from it
    2. They understand only English and cannot write Hindi
    3. They generate text by predicting likely next tokens, so fluent output can still be false
    4. They cannot write computer code
    Show answer

    Answer: C. An LLM builds text token by token from learned patterns; it has no built-in truth check, so a fluent answer can be wrong. LLMs commonly handle many languages and can write code.

  2. 2. In "Generative Pre-trained Transformer", what does "pre-trained" mean?

    1. It was first trained on a very large general text collection before being adapted for specific uses
    2. It was trained only on questions from previous exams
    3. It was trained by the user before first use
    4. It needs no training at all
    Show answer

    Answer: A. Pre-training is the first stage, on a huge general text collection; fine-tuning and human feedback then adapt the model. Users do not train it, and it is not limited to exam questions.

  3. 3. An app turns the prompt "a tiger in a mustard field at sunset" into a new picture. Which kind of system is it most likely to be?

    1. A spam classifier
    2. A search engine retrieving an existing photo
    3. A speech-recognition system
    4. A diffusion-based image generator
    Show answer

    Answer: D. Creating a new image from a text description is generative AI, commonly done by diffusion models that remove noise step by step guided by the prompt. A search engine finds existing photos instead of making one.

  4. 4. "You are a geography teacher. Explain monsoon winds. I am a beginner at Class 10 level." Which part of a strong prompt is missing?

    1. Role
    2. Format - how the answer should be laid out
    3. Task
    4. Context
    Show answer

    Answer: B. Role (geography teacher), task (explain monsoon winds) and context (Class 10 beginner) are present. Nothing says how to lay out the answer - points, a table or a word limit - so format is missing.

  5. 5. After reading an AI explanation of the Preamble, which follow-up best turns it into self-test material?

    1. Thanks, that was helpful
    2. Explain it again in the same way
    3. Make the explanation longer
    4. Make 5 MCQs on this with four options each, and list the answers with one-line reasons only at the end
    Show answer

    Answer: D. A quiz with answers held back until the end lets you test recall honestly. Thanking, repeating or lengthening the explanation does not test you.

  6. 6. Which change most improves the weak prompt "write about rivers"?

    1. Naming the task, scope, audience and format, such as a revision table of six peninsular rivers with set columns
    2. Adding "please" at the end
    3. Typing it in capital letters
    4. Repeating the word "rivers" three times
    Show answer

    Answer: A. The prompt lacks scope, purpose and format; adding them gives a usable answer. Politeness, capital letters or repetition add no information.

  7. 7. Which piece of AI output needs the most careful checking before you rely on it?

    1. A suggested daily study timetable
    2. An analogy explaining photosynthesis
    3. A statement of the application fee and age limit for a particular recruitment
    4. A rewording of your own notes
    Show answer

    Answer: C. Fees and age limits are official facts that change between notifications and may be after the training cut-off; only the official notification settles them. The other outputs are study aids you can judge yourself.

  8. 8. A student uses AI to translate English notes on the Fundamental Duties into Hindi, and one legal term comes out in an unusual form. What is the best action?

    1. Keep the AI's version, since translation tools are always accurate
    2. Compare the term with a standard Hindi-medium textbook or the official Hindi text
    3. Delete the Hindi notes and study only in English
    4. Ask the same tool to translate it again and accept whichever version appears twice
    Show answer

    Answer: B. Legal and constitutional terms have fixed standard forms, which a standard textbook or official text shows. Repeating the request only gives the same tool's guess again.

  9. 9. A student has 25 days, studies 2 hours a day and keeps 5 days only for mock tests. The chatbot's plan gives 48 hours of topic study. What should the student conclude?

    1. The plan is not realistic: only 40 hours are available, so ask the chatbot to rebalance it
    2. The plan is realistic because it is close to 50 hours
    3. The plan is realistic because the chatbot calculated it
    4. The plan is not realistic because 60 hours are available
    Show answer

    Answer: A. Study days = 25 − 5 = 20; hours = 20 × 2 = 40. The plan asks for 48, which is more than 40, so it cannot be followed and should be rebalanced.

  10. 10. Asked for references, a chatbot lists three book titles with authors, but none can be found in any library or publisher catalogue. This is best described as:

    1. The training cut-off
    2. Bias in the training data
    3. A translation error
    4. Hallucination - invented references
    Show answer

    Answer: D. Producing plausible-looking but non-existent references is hallucination. A cut-off would explain missing recent information, not invented books.

  11. 11. Which pair contains items that should both never be pasted into a chatbot?

    1. A textbook paragraph and a maths question
    2. An OTP and an Aadhaar number
    3. Your own essay draft and a list of syllabus topics
    4. A public news headline and a list of chapter names
    Show answer

    Answer: B. OTPs and Aadhaar numbers are secret or personal identifiers that can be misused if stored or exposed. The other pairs contain no secrets or identifiers.

  12. 12. Which is the sound way to confirm an important fact a chatbot gave you?

    1. Ask the same chatbot again and accept the answer if it repeats it
    2. Accept the fact if two different chatbots agree
    3. Check the official source or a standard textbook, and confirm in at least two independent reliable sources
    4. Accept the fact if the answer is long and detailed
    Show answer

    Answer: C. Official sources and textbooks are authoritative, and two independent reliable sources guard against one source's error. Chatbots can repeat or share the same mistake, and length is no sign of accuracy.

  13. 13. A viral video shows a well-known official announcing that an exam is postponed; the lip movements look slightly out of step. What is the best action?

    1. Forward it quickly to warn your friends
    2. Assume it is true because the face is familiar
    3. Stop studying until the confusion clears
    4. Check the exam body's official website or press release before believing or sharing it
    Show answer

    Answer: D. Out-of-step lips are a possible deepfake sign, and only an official notice settles the matter. Forwarding an unconfirmed clip spreads possible false news.

  14. 14. Which combination is the clearest warning sign of a voice-clone fraud call?

    1. Urgency, a demand for money or an OTP, and pressure to keep it secret
    2. The caller speaks in a familiar voice
    3. The call comes in the evening
    4. The caller asks how your studies are going
    Show answer

    Answer: A. Fraud calls rely on urgency, a request for money or codes, and secrecy so you do not check. A familiar voice is what cloning imitates, so on its own it proves nothing either way.

  15. 15. Which of these is plagiarism?

    1. Using AI to make your own revision timetable
    2. Submitting a chatbot-written essay as entirely your own work
    3. Asking AI to quiz you on a chapter
    4. Writing your own essay and stating that AI suggested an outline, as the rules allow
    Show answer

    Answer: B. Presenting AI-written text as your own work is plagiarism. Making a timetable, being quizzed or declaring permitted AI help are honest uses.

Where to revise: questions 1-3 - lesson 1 (what generative AI is); questions 4-6 - lesson 2 (writing good prompts); questions 7-9 - lesson 3 (using AI to study honestly); questions 10-12 - lesson 4 (limits, checking and data safety); questions 13-15 - lesson 5 (deepfakes, voice clones and plagiarism).

Sign in to keep track of the lessons you finish.

Privacy · Terms · Refunds · Delivery