Module 10 – Responsible AI
beginner20 XP

When AI Gets It Wrong! 😅

The hilariously wrong and seriously scary mistakes AI makes — and how to spot them

When AI Gets It Wrong! 😅

AI is Smart But Not Perfect

Even the best AI in the world makes mistakes. Some are funny, some are serious — all of them teach us something important!

😂 Hilariously Wrong AI (Real Examples!)

The Cheese Burger vs Cat Problem

A major facial recognition system was tested by a researcher who found it could be fooled by holding up a photo of a cheeseburger in front of their face. The AI thought it was a human! 🍔

The Microwave Incident

Someone asked an AI assistant: "How do I stop my microwave beeping?" AI confidently replied: "Simply remove the beeping module from the control board by unscrewing the back panel..."

There is no "beeping module." The AI made it up completely! 😱 (Just press the mute button!)

The Wrong Language

An AI translating customer complaints for a company translated everything correctly EXCEPT the angry words — it replaced all swearing with polite alternatives, so the company never knew how upset customers really were! 😂

😰 Seriously Wrong AI (Real Consequences)

The Lawyer Who Got Caught

In 2023, a real American lawyer asked ChatGPT to help find legal cases to support his argument.

ChatGPT gave him 6 convincing court cases with names, dates, and details.

Every single case was completely made up.

The lawyer submitted them to a judge. The judge was NOT happy. The lawyer was fined and nearly lost his licence! 😬

The Fake Photo Problem

AI image generators can create incredibly realistic photos of things that never happened — fake news photos, fake evidence, fake events.

🧠 The "Hallucination" Explained

When AI "makes things up confidently," it's called a hallucination:

AI generates text by predicting: "What word is MOST LIKELY to come next?" Problem: "Likely" ≠ "True"! AI can generate convincing-sounding facts that are completely wrong.

✅ How to Stay Safe

  1. Never use AI for important facts without checking
  2. Always verify medical, legal, or safety information from professionals
  3. Check dates — AI knowledge has a cutoff
  4. Be sceptical of specific numbers (AI loves inventing statistics)
  5. Trust your gut — if something sounds odd, check it!

Quick check

01/3

What is an AI 'hallucination' and why is it dangerous?