AI for Kids Worksheet: Part 1 Test
AI tip (Name the idea, then answer): Learning from examples → AI; fixed rules → a normal program. Order: training data → model → prediction; good data is many, varied and correctly labelled. Generative AI makes new things and can hallucinate; one-sided data causes bias.
NameDateTime takenScore ___ / 40
- 1.🔍 Which is the odd one out? (a) blurry, unclear photos (b) many varied examples (c) clear, sharp examples (d) correct labels
- 2.Which of these is generative AI (it makes something new)? (a) predicting rain (b) suggesting a video (c) inventing a recipe (d) spotting spam
- 3.📚 Good or bad for training a model: many varied examples? (a) good for training (b) bad for training
- 4.🖼️ If an image generator always draws a scientist as a man, that shows… (a) a strong battery (b) bias from data (c) a broken screen (d) a fair model
- 5.🛠️ What helps make an AI system fairer? (a) one group only (b) fewer tests (c) ignoring mistakes (d) more varied data
- 6.True or false: “learns from examples” describes how a normal program works. (a) True (b) False
- 7.True or false: a voice app trained only on adult voices is balanced data (fairer for everyone). (a) True (b) False
- 8.🔍 Which is the odd one out? (a) AI can be wrong (b) AI can make up facts (c) AI knows everything (d) AI copies data patterns
- 9.Which of these is a prediction? (a) labelled fruit photos (b) flagging a new email (c) past weather records (d) 500 labelled leaf photos
- 10.Which of these is training data? (a) flagging a new email (b) naming a new leaf (c) the learned patterns (d) old emails marked spam
- 11.🧐 Fact or myth: “AI learns from data”? (a) a fact (b) a myth
- 12.True or false: “naming a new fruit photo” is the model. (a) False (b) True
- 13.True or false: “check who is missing” adds to bias. (a) False (b) True
- 14.Which of these is good for training a model? (a) copies of one photo (b) only one kind of example (c) blurry, unclear photos (d) checking labels twice
- 15.Which of these is the model? (a) recorded spoken words (b) labelled fruit photos (c) naming a new leaf (d) what training builds
- 16.⏩ In machine learning, which comes first: typing a new spoken word or 500 labelled leaf photos? (a) typing a new spoken word (b) 500 labelled leaf photos
- 17.🏷️ You are labelling messages to train a spam filter. How should this one be labelled? “Last chance! Free gift card inside” (a) spam (b) not spam
- 18.🛠️ Does this reduce bias or add to it: use one group only? (a) adds to bias (b) reduces bias
- 19.Which of these describes how a normal program works? (a) gives a likely answer (b) improves with more data (c) spots patterns in data (d) never learns from data
- 20.True or false: voices of children and adults from many regions is balanced data (fairer for everyone). (a) True (b) False
- 21.Which of these is the model? (a) naming a new leaf (b) the learned patterns (c) tomorrow’s rain guess (d) recorded spoken words
- 22.Which of these is bad for training a model? (a) photos in many lights (b) many varied examples (c) blurry, unclear photos (d) new data for testing
- 23.🤖 AI or a normal program: which one “learns from examples”? (a) a normal program (b) AI
- 24.🏷️ What is a “label” in machine learning? (a) the price (b) the right answer tag (c) a sticker on a laptop (d) the font size
- 25.🔁 Training data, the model or a prediction: “recorded spoken words”? (a) the model (b) a prediction (c) training data
- 26.True or false: “does exactly as coded” describes how a normal program works. (a) False (b) True
- 27.⚖️ One-sided or balanced data: leaves from every season of the year? (a) one-sided data (b) balanced data
- 28.📚 An AI chatbot names a book that does not exist. This is called… (a) a summary (b) a hallucination (c) a password (d) a backup
- 29.🔍 Which is the odd one out? (a) add more of the same (b) test on one group only (c) ask different people (d) copy old unfair choices
- 30.🤖 What makes an AI system different from a normal program? (a) it has a screen (b) it uses electricity (c) it is always right (d) it learns from examples
- 31.🔦 Is a torch with an on/off switch AI? (a) only if it is bright (b) yes (c) no
- 32.True or false: “a picture of a castle made of mangoes” asks for an image. (a) True (b) False
- 33.✅ Is this a real check of an AI answer: ask your teacher? (a) not a real check (b) a real check
- 34.📚 Good or bad for training a model: correct labels? (a) bad for training (b) good for training
- 35.Which of these is AI that sorts, spots or predicts (not generative)? (a) writing a new poem (b) making up a story (c) suggesting a video (d) creating a cartoon
- 36.True or false: “old emails marked spam” is a prediction. (a) False (b) True
- 37.Which of these adds to bias? (a) balance the data (b) add varied examples (c) test on one group only (d) test on many groups
- 38.Which of these is generative AI (it makes something new)? (a) unlocking with a face (b) sorting photos by face (c) suggesting a video (d) drafting a letter
- 39.Which of these uses AI? (a) a light switch (b) a digital watch (c) face unlock on a phone (d) a printed timetable
- 40.True or false: “recorded spoken words” is the model. (a) False (b) True
Answer key
- blurry, unclear photos
- inventing a recipe
- good for training
- bias from data
- more varied data
- False
- False
- AI knows everything
- flagging a new email
- old emails marked spam
- a fact
- False
- False
- checking labels twice
- what training builds
- 500 labelled leaf photos
- spam
- adds to bias
- never learns from data
- True
- the learned patterns
- blurry, unclear photos
- AI
- the right answer tag
- training data
- True
- balanced data
- a hallucination
- ask different people
- it learns from examples
- no
- True
- a real check
- good for training
- suggesting a video
- False
- test on one group only
- drafting a letter
- face unlock on a phone
- False
Free ai for students lessons, practice and worksheets at talentjr.in/ai-for-students