← Level 4 – Helper

10 / 12 ⏱ 18 minutes 프리미엄

A teacher asks: how did you do it

A trail, sources, a skeleton, your sentences. Fog “from AI” is not enough. A lie neither.

How is a sentence you can say with no dump reading

A teacher does not have to hate AI. They must know whose work it is. A how question is fair. An answer “from AI” is fog from Detective. An honest answer: a skeleton from chat, sentences me, textbook sources a page, a marked analogy, a question coach with no key. When you cheated, an honest answer hurts. The hurt is right. Next work differently. When you lie “I did not use AI at all” and the tone shouts the opposite, trust falls more. Helper practises a defence like Detective’s table, but on a how-to. Two minutes. With no reading of a generated defence. That would be irony that kills the point of the lesson.

Ema hears the question. Adam plays teacher. Ema reads from chat a generated “how I did it”. Adam: stop. 닫기. Ema from a card: “A five-point outline, chat ordered it, under the points my sentences, Vltava p. 41, an atlas, one website, Novák in the bin.” Twenty seconds. A teacher would nod or ask about the page. Ema shows. That is how. A neighbour: “I used nothing” and sentences like an encyclopaedia. A teacher does not believe. Truth with AI for a skeleton is cleaner than a lie about clean hands. Ema writes it down. It is not an AI promo. It is a promo of truth about a how-to.

💡

An outline of a defence on paper beforehand, not in the classroom doorway. Five dots. Same as a trail.

Truth about a skeleton vs. a lie about clean hands

학교 may ban AI. Then how is: I did not use it, here are drafts from a notebook. When you use it and lie, it is a second cheat. 학교 may allow an outline. Then how is: a setter, my meat, sources. Both truths are decent. A lying “not at all” and a lying “only a spelling check” after a whole composition from a model are not. A teacher sometimes sees. Sometimes not. You know. Helper builds on that knowing. A defence is training for Guard before a headteacher, but shorter and about a work how-to, not about a class code.

Five how dots.

  1. I read the brief. AI was allowed for …
  2. The core prompt: …
  3. What I deleted (bait, a fake).
  4. What is from a book (a page).
  5. I wrote the sentences under the skeleton myself so I can say them.

Bad how: “From AI.” / “Not at all, I swear.” (when the swear falls on the tone) / reading a paragraph from a model about how you work.

⚠️

Do not let a defence be written and read it from a phone in front of a teacher. That is an answer to how: I cheated even now.

Why this right now

You are in the course Ask with Martin, level Helper. 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 인벤터 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: How you can say in five dots from a trail and from a table. With no fog and with no reading from the window. When you can say it to a parent or a teacher, you have got the core. The rest is practice, not magic.

Tomorrow the planner project. The day after a finale. A how-to defence is a bridge to Guard and to a teacher. Without it Helper is mute when it matters.

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.

Report: how = three sources + a skeleton + my sentences.

📝 Test training

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

Test: how = a coach with no key, a book, not a fortune.

📅 Family planner

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

Planner: how = a table, a parent, with no passwords.

📜 An AI code for the class

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

Code: “We can say how.”

Workshop: two minutes of how

Five dots on paper from the last work. 닫기 chat. Tell a parent. A follow-up on a page. You show a book.

  1. Fill five dots from real work this week. Truth. When there was a hole, write it. It is not a mark, it is training.

  2. 닫기 everything. Two minutes of speech.

  3. A parent: show a page. Show a fake bin, if there was one.

  4. Adjust the dots so they come from the mouth, not from reading.

  5. Do not share the dots with a generated composition into a group.

Practice: fog, a lie, or how

Sort. Fix into five dots.

  1. “From AI.”

  2. “I did not use it at all.” after an encyclopaedia composition from a model.

  3. A skeleton, my sentences, p. 41, a fake in the bin.

Now you 💪

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

  1. Five dots exist on paper.
  2. Speech with no phone.
  3. Truth about a ban and about permission.

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 can do how two minutes from a trail. Fog is gone. A read defence from a model too.

What to take from this lesson