AI is powerful but not perfect — a smart explorer knows its limits. AI chatbots can "hallucinate": make up facts, fake quotes, or wrong answers that sound totally believable. AI can also be biased (unfair) if it learned from biased data, and it can be confidently wrong about maths, dates, or recent events. The rule for using AI wisely: treat it like a helpful but sometimes-mistaken assistant, not an all-knowing oracle. Use it to brainstorm, explain, and draft — but check anything important against a trusted source, especially facts, numbers, and news. The human stays in charge.
Technique explanation
Worked examples
🧩 Unplugged activity: Think of a question where being wrong would matter a lot (a medicine dose, a history date for a test, directions). Write down where you would double-check the answer instead of trusting AI alone.