AI School › AI for studying · Lesson 49 of 130

⚠️ The limits: sometimes AI is wrong

AI doesn't know everything and can say false things very confidently: knowing its limits protects you.

Even the car's navigator, usually excellent, sometimes sends you down a closed road or into a field. Do you trust it blindly or also look out of the window? AI is like that navigator: very useful, but not perfect. It can get dates wrong, mix up people, invent a book that doesn't exist, botch a calculation. And it does so with the same confident voice as when it's right, and that's exactly the danger: it always sounds sure, even when it has made a huge blunder.

Why does it happen? Because AI doesn't really know things the way you do: it puts together words that usually sit near each other, and sometimes the combination is wrong. It has no eyes to see the world, nor does it know what happened after the day it was trained. So what do you do? You treat information like mushrooms in the woods: beautiful, but before eating them you check with an expert. For important facts, always verify in a book or a safe source. Knowing AI can be wrong doesn't make you use it less: it makes you use it better, as a smart person.

🎒 Try it in class (or at home)

Try to catch the AI out: ask it something very specific or made up and see if it admits it doesn't know or invents an answer.

← Lesson 48: AI as a patient tutor · Lesson 50: Studying with your own brain switched on →

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.