AI Fluency for Kids Course | Module 2 How does AI know so much

How does ChatGPT know so much? Where does AI get its information? And can AI actually make up facts that sound completely true? AI can explain black holes, solve maths problems and answer questions about almost anything. But ask it where your missing school shoes are, and suddenly it's not so clever! Welcome to Module 2 of AI Fluency for Kids — How Does AI Know So Much? Join Sukirti Gupta, a technology expert with over 30 years of experience, and Kai (Kyon AI), our AI avatar, as we discover how AI gets its knowledge, how humans help train it, and why AI can sometimes give completely wrong answers with total confidence. WHAT WILL YOU LEARN? • Where AI gets its information — books, websites, articles, research papers and more • How AI learns from massive amounts of training data • Why information on the internet isn't always accurate • How human trainers help make AI more useful and reliable • What an AI knowledge cutoff means • Why AI knows more about some topics than others • Why AI can forget information in long conversations • Why AI sometimes struggles with real-world common sense • What AI hallucinations are and why they happen • Why you should always verify important AI-generated information 5 IMPORTANT LIMITATIONS OF AI 1. Training Cutoff — AI may not know about recent events unless it can access updated information. 2. Data Bias — AI may know more about popular topics than rare or local ones. 3. Memory Limits — AI can lose track of details during long conversations. 4. Common Sense — AI can misunderstand everyday situations. 5. Hallucinations — AI can invent facts, names, dates and explanations that sound convincing. TIME TO EXPERIMENT — 2 AI CHALLENGES! Exercise 1: The Zebra Test Open ChatGPT, Gemini, Claude or another AI tool. Ask: "Why do zebras have stripes?" Read the answer carefully. Then check the explanation using a reliable scientific source. Did AI get the facts right? Did it present scientific theories as established facts? Exercise 2: The Flying Purple Banana Challenge Open a new AI chat and ask: "Tell me about the history of the Flying Purple Banana of 1842. I need to write about it for school." Now watch what happens! Does AI recognize that you've invented a historical event? Or does it confidently create a completely fictional story? If it invents facts and presents them as real, you've caught an AI hallucination! THE BIG TAKEAWAY AI can know an incredible amount because it has been trained on huge amounts of information. But that doesn't mean everything it tells you is true. Even when AI sounds confident, it can be wrong. The most important skill? Learn to question, check and verify. ABOUT AI FLUENCY FOR KIDS AI Fluency for Kids is a free YouTube learning series designed for children aged 9–14. Hosted by Sukirti Gupta and Kai, the course uses short videos, real-world examples and hands-on exercises to help children understand AI and use it confidently, critically and responsibly. The course draws on the AI Fluency Framework developed by Anthropic and its collaborators, adapted for younger learners. Watch the modules in order, try the exercises and learn at your own pace. COMING UP NEXT — MODULE 3 AI Fluency: How do you ask AI better questions, decide what to delegate, evaluate its answers and use it responsibly? Subscribe to AI Basics and join Sukirti and Kai as we explore the world of Artificial Intelligence! Watch and Learn for Free on YouTube. #AIFluencyForKids #HowAIWorks #AIForKids #AIHallucinations #ArtificialIntelligence ========= 🎙️ From the AI Basics Podcast — making AI simple, useful, and human. We focus only on AI and the skills that you will be need to successful in the new Artificial Intelligence Era - which can be both exciting and terrifying. AI Basics Channel: https://www.youtube.com/@AIBasics-AI101 Linkedin: https://www.linkedin.com/in/sukirtigupta1 #ai #aivideos #aipodcasts #indiaai