Ask again differently · Code with Martin

When the answer does not sit, you do not change the brand. You change the angle of the question.

A second angle, not rage at a logo

The first answer is a draft. Sometimes it misses. People then shout that AI is stupid, and open another brand with the same mess. Curious stays and changes the angle. Instead of “explain again” — which often dumps the same thing longer — you say: an analogy with a kitchen, an analogy with sport, only steps, no foreign words, the opposite: what it is not. That is round two. Round three is a narrowing: only one example. You stay in one chat, so the model sees what already landed. When after three angles it still misses, you go to the textbook. That is not a loss. That is a referee. Another brand with the same angle often repeats the same trap.

Ema gets a lecture about air pressure she does not understand. Adam advises: “Put Gemini, that is cleverer.” Ema tries in the same window: “Explain it like pumping up a ball. No formulas.” Suddenly she sees it. Adam tries a third angle: “Say what it is not.” The model says it is not a water pump in a tap. Adam gets the mistake he had in his head. The textbook confirms it with a picture. No new logo. Three angles. Ema writes a list: analogy, steps, what it is not. Next time, when she is angry, she will not open an app shop. She will open the list.

💡

A ban on the sentence “explain it better”. It is empty. Say how else: an analogy, steps, the opposite, one number.

Angle, narrowing, referee

An angle changes the picture: kitchen, pitch, room. Narrowing takes width: one example, five sentences, no dates. The referee is a source outside the chat. Three tries are fair. A thirtieth is wandering that yesterday’s timer solved. A new chat wipes context — sometimes you want that, when the train raced into personal stories. For schoolwork it is better to stay and point: “the previous analogy with a ball was good, now only what it is not.” The model then does not build from zero. Again differently is a craft, not hysteria. Hysteria changes the brand. Craft changes the angle.

A round-two menu. Pick one, not all at once.

“Same topic, an analogy with a kitchen. No formulas.”
“Only three numbered steps. No intro.”
“Say what it is not. One mistake people make.”
“Only one number example, do not say the answer, I will calculate it.”
“Cut it to five sentences. Explain foreign words in brackets.”

A bad round two: “You are useless, try again.” That is rage. Rage will not bring the model closer to truth.

⚠️

A third angle that still asks for finished homework is still a foul. A cheat that sounds different is a cheat.

Why this right now

You are in the course Ask with Martin, level Curious. You work in a chatbot (ChatGPT, Gemini or Claude) — same game, different brand. No code. A parent or the school creates the account. A child’s email does not belong here.

This is not the Inventor level. There you build a chatbot on your own website. Here you are a user: a question, a check, a citation.

You will know this lesson by one sentence: It does not sit? Change the angle, not the club. Then open a book. When you can say it to a parent or a teacher, you have got the core. The rest is practice, not magic.

You can do a shorter prompt. Now you can do a second try without rage. Tomorrow you add age eleven — that is really a named angle. Today you have a menu.

Four things you will build in this course

The course has four small projects. You do not have to do them all at once. Today you add only what this lesson can do.

🏫 A report with citations

An outline, three sources that exist, and a sentence in your own words. AI must not be the only author.

Outline misses? Ask again: only five titles, no paragraphs.

📝 Test training

Coach role: let it ask, do not let it say the answer. You count / answer yourself.

Coach dumps answers? Angle: only questions, no numbers at the end.

📅 Family planner

Tasks for the week (not passwords, not marks). A parent sees the plan, not your chat.

Planner is stuffed? Again: only three tasks a day.

📜 An AI code for the class

Seven short rules on a poster. No horror — a how-to.

Code: “First another angle, then another brand.”

Workshop: three angles on one bit of work

One topic from a textbook. Three messages. Three angles from the menu. Then a book as referee.

Pick a term you do not fully get. First prompt with a stamp: explain with one analogy. Send it.
Second message: another angle from the menu (steps, or what it is not). Do not open a new brand.
Third message: a narrowing to five sentences. If you already get it, the third can be: “ask me one question so I can check myself.”
Open a textbook. Which angle sat? Which hallucinated a detail? Cross out the detail.
In the notebook copy the menu of five angles. Do not have the sentence “explain it better” there.

Practice: rewrite rage into an angle

Three angry sentences. Swap them for a concrete angle.

“You are stupid, say it again.”
“This is too long, so somehow…” (narrowing: how many sentences?)
“I will put it in another bot.” What do you try before that?

Now you 💪

Go through three checks. At each one tick in your head: can I see it, or not?

You have a menu of angles in the notebook.
You will not say an empty “explain it better”.
After three angles you go to a book, not to a thirtieth round.

If something is missing, go back to the steps above. This is not a test for marks. It is a safety net, so you go on sure, not confused.

✅

Check: You turned one bit of work with three angles and the referee was the textbook.

What to take from this lesson

Round two = another angle, not another club.
A ban on an empty “explain it better”.
Narrowing is an angle too.
After three tries a referee outside the chat.
Rage at a model will not add truth.
The names in the examples (Ema, Adam) are made up. Do not put your number, email or photo here.
Next lesson: When you can do this, go to the next lesson in the same level.