Which answer is right?

5 minutes. One machine makes one widget in 5 minutes. So 100 machines make 100 widgets in the same 5 minutes. Autopilot sees the pattern “5-5-5, so 100-100-100”, and that’s the trap. It’s the second question from the same Frederick test.

When should you call the pilot?

Before putting $300 into a coin you heard about an hour ago. All three signals at once: high stakes ($300 of $500), a new situation and it’s about odds. Perfect time to ask: “What do I know about this coin, and how did I check it?”

Why does half the audience miss the gorilla?

Attention is busy counting, so the brain filters out the rest. Attention is limited. When the pilot is loaded with a task, autopilot misses even the obvious. Without the counting task, nearly everyone sees the gorilla.

What’s the best thing to do with autopilot?

Trust it in routine and check it in risky situations. Autopilot is usually right and saves energy. It’s dangerous only in narrow situations: high stakes, novelty, numbers. That’s where you bring in the pilot.

Autopilot and pilot

Lesson 1 of 31 in the course “Brain Hacks”. Free, with animations and a quiz.

Two pilots in your head

Meet Danya, 18. He’s saving his first $500, studying and doomscrolling TikTok. He sees an angry face and instantly knows: that person is mad. That’s autopilot: fast, automatic, effortless. Now work out 17 × 24. The answer doesn’t just pop up; you have to focus and go step by step. That’s the pilot. Psychologists call them “System 1” and “System 2”. In this course they’re simply autopilot and pilot.

Test: the bat and the ball

Answer fast. A bat and a ball cost $1.10 together. The bat costs $1 more than the ball. How much is the ball? “10 cents” pops into your head. Check it: then the bat is $1.10 and the total is $1.20. The right answer is 5 cents: ball 5¢, bat $1.05, total $1.10. Psychologist Shane Frederick designed the puzzle to catch the moment autopilot answers for you.

Even top students get it wrong

Frederick gave this test to thousands of students. At Harvard, MIT and Princeton more than half answered “10 cents”. At less selective universities over 80% got it wrong. It’s not about being smart. The puzzle is easy; the answer just looks plausible, and the pilot is too lazy to check it. Smart people fail in exactly the same way when they skip the check.

The invisible gorilla

Psychologists Daniel Simons and Christopher Chabris filmed two teams passing a ball and asked viewers to count the white team’s passes. Midway, a person in a gorilla suit walks across, thumps their chest and leaves. They’re on screen for 9 seconds, yet about half the viewers don’t notice. Attention is limited: when the pilot is busy, autopilot filters out everything else. We’re blind to the obvious, and blind to our own blindness.

You know it, but you still see it

Look at the two lines. The bottom one looks longer, but measure them and they’re identical. That’s the Müller-Lyer illusion. The fun part: even when you know the answer, your eyes keep seeing a difference. You can’t switch autopilot off by willpower. You can only learn the trap and stop trusting the feeling in those situations. Thinking errors work the same way, and this whole course is about spotting them.

Autopilot isn’t the enemy

Autopilot runs almost your whole day: reading signs, doing 2 + 2, spotting emotions, walking a familiar route. And it’s usually right. The pilot is slow and tires fast, but only it can compare options, calculate and check logic. If the pilot checked every step, you wouldn’t even make it through breakfast. The goal isn’t to turn autopilot off; it’s to know when not to trust it.

When to call the pilot

Three signals that autopilot can’t be trusted right now. High stakes: money, health, relationships. A new situation: autopilot has no experience and is guessing. Numbers and odds: that’s where autopilot is weakest, like in the bat puzzle. See even one signal? Ask the stop question: “What answer came first, and how did I check it?” The rest of the time, live on autopilot in peace.

Try it: a decision diary

For 3 days in a row, write down one decision a day. Three columns: what you decided, what autopilot suggested, whether the pilot checked. Danya’s example is in the picture: he checked the 30%-off sneakers by finding the pre-sale price. He checked the TikTok coin by finding out who made it. He didn’t check his route to school, and that’s fine. The goal isn’t to check everything; it’s to notice where autopilot decides for you. That’s the foundation of the whole course.

Check yourself

Which answer is right?
Correct answer: 5 minutes. One machine makes one widget in 5 minutes. So 100 machines make 100 widgets in the same 5 minutes. Autopilot sees the pattern “5-5-5, so 100-100-100”, and that’s the trap. It’s the second question from the same Frederick test.
When should you call the pilot?
Correct answer: Before putting $300 into a coin you heard about an hour ago. All three signals at once: high stakes ($300 of $500), a new situation and it’s about odds. Perfect time to ask: “What do I know about this coin, and how did I check it?”
Why does half the audience miss the gorilla?
Correct answer: Attention is busy counting, so the brain filters out the rest. Attention is limited. When the pilot is loaded with a task, autopilot misses even the obvious. Without the counting task, nearly everyone sees the gorilla.
What’s the best thing to do with autopilot?
Correct answer: Trust it in routine and check it in risky situations. Autopilot is usually right and saves energy. It’s dangerous only in narrow situations: high stakes, novelty, numbers. That’s where you bring in the pilot.

The lazy pilot →