Lesson 22 of 1234
AI Can Be Totally, Confidently Wrong
AI sounds sure of itself even when it is making stuff up. Here is how to notice when it is wrong and what to do about it.
Lesson map
What this lesson covers
Learning path
The main moves in order
- 1The Confident Guesser
- 2hallucination
- 3fact checking
- 4confidence
Concept cluster
Terms to connect while reading
Section 1
The Confident Guesser
Here is a weird thing about AI chatbots. They always sound sure of themselves. Even when they are completely wrong. Even when they are making stuff up.
Scientists call this a hallucination, which is a fancy word for when an AI invents a fact that is not true. The AI is not lying on purpose. It does not know it is wrong. It is just guessing the next most likely thing to say, and sometimes the guess is fake.
When AI is most likely to be wrong
- Exact numbers (dates, populations, sports scores)
- Names and who said what
- Very new things (the AI may not know about last week)
- Things only a few people have written about online
- Lists that feel too neat or too long
Warning signs
- The answer feels too perfect or too fast
- The AI names a book, article, or website you cannot find
- Two AIs give you different answers to the same question
- The AI gets super specific with a number it has no way to know
Your fact-checking toolkit
- 1Look up the fact on a site you trust (Wikipedia, a real news site, a real book)
- 2Ask the AI how sure it is, or to cite its source
- 3Ask a second AI the same question and see if they match
- 4If it is important, ask a real human who knows the topic
“AI is the most confident liar I have ever met, and it does not even know it is lying.”
Key terms in this lesson
The big idea: AI sounds sure even when it is wrong. Always check the important stuff. Treat AI like a smart friend who makes things up sometimes.
End-of-lesson quiz
Check what stuck
15 questions · Score saves to your progress.
Tutor
Curious about “AI Can Be Totally, Confidently Wrong”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Explorers · 15 min
The Golden Rule, But With AI
You can do things with AI you could never do before. That means you can also hurt people in new ways. Here is the simple rule that keeps you on the right side of the line.
Explorers · 20 min
Real or Fake? Spotting AI Pictures and Videos
AI can now make pictures and videos that look absolutely real. Here are the signs to look for and the habits that will keep you smart.
Explorers · 20 min
Deepfakes: When a Fake Looks Like Someone You Know
A deepfake is a fake video or voice that looks and sounds like a real person. Here is what they are, why they hurt people, and what to do if you see one.