Start learning →

Claude AI AI Basics — practice questions

Explore 10 free AI Basics questions in Claude AI Claudearchitect. Open each question or exercise to read its content, reference answers and available explanation.

Practise AI Basics with Quill →

AI Basics questions and exercises with answers

Showing questions 1–10 of 10.

Which stage is MISSING from an AI language model's pipeline?Read the answer and worked solution →During fine-tuning, what do human raters actually do?Read the answer and worked solution →Why is AI especially strong at language tasks like summarising and rewriting?Read the answer and worked solution →Which task is AI best suited for?Read the answer and worked solution →What does AI do well when you ask it to compare two options?Read the answer and worked solution →AI says an essay has 850 words. The real count is 1003. What does this show?Read the answer and worked solution →What is the main reason AI sometimes produces confident but wrong answers?Read the answer and worked solution →Which type of claim carries the highest hallucination risk?Read the answer and worked solution →A friend asks: what should you do first when an AI answer contains a specific date?Read the answer and worked solution →Pop the jobs a model does wellRead the answer and worked solution →

Worked practice questions

Which stage is MISSING from an AI language model's pipeline?

  1. Reading the input prompt from the user
  2. Matching patterns from training data to the input
  3. Outputting the predicted text back to the user
  4. Checking whether the predicted output is factually true
Show answer and solution

Correct answer: D. Checking whether the predicted output is factually true

Card 4: AI predicts plausible text but has no built-in truth-verification stage. Source: simple_wikipedia_ai_ocr.md Limits section -- AI cannot do things in the same way as people do, and lacks common sense.

Open this question →

During fine-tuning, what do human raters actually do?

  1. Write new training data from scratch
  2. Add more internet text to the training set
  3. Show the model examples of good and bad replies
  4. Fix bugs in the model's code directly
Show answer and solution

Correct answer: C. Show the model examples of good and bad replies

Card 3 of this block: in fine-tuning, humans show the model pairs of good replies vs bad replies so the model learns the shape of being helpful. Source: claude_constitution_ocr.md (constitution plays a crucial role in the training process, p.1).

Open this question →

Why is AI especially strong at language tasks like summarising and rewriting?

  1. It connects to the internet to find answers
  2. It was trained on huge amounts of written text
  3. It stores a copy of every book ever written
  4. It uses a built-in grammar rule library
Show answer and solution

Correct answer: B. It was trained on huge amounts of written text

Card 1 in this block: AI's training was on vast text data so language is its strongest skill. Source: claude_constitution_ocr.md line 108 (coding, writing, analysis as everyday tasks) and simple_wikipedia_ai_ocr.md line 12 (natural language processing listed as a core AI application).

Open this question →

How to practise

  1. Choose a AI Basics question and try to work out your answer.
  2. Open the question to check the correct answer and read its worked solution.
  3. Continue in the app for bite-sized lessons, explanations and review with Quill. You can start without creating an account.

More Claudearchitect topics

Claudeai PromptingClaudeai Long ContextClaudeai McpClaudeai Model PickingClaudeai RagClaudeai AgentsClaudeai CachingClaudeai Code