Helpful Assistance vs Outsourcing the Thinking

By Better Wise

You have drafted, argued, and checked AI suggestions. Now comes a quieter skill: telling the difference between AI that sharpens your thinking and AI that quietly replaces it. This matters for your capstone because the final essay must still be yours, and your reflection must describe real decisions you made. In the last lesson you verified suggestions before they entered your draft. This lesson gives you a short test for deciding which suggestions should enter at all.

By the end of this lesson, you can:

• Explain why sequence and cognitive work matter more than tool use.
• Apply a four-question decision test to a real AI exchange.
• Classify gray-zone uses such as paraphrasing, brainstorming, and outlining.
• Write a short usage note that makes your choices visible.

The line is about thinking, not tool use

A common mistake is to ask, "Is using AI cheating?" That question is too broad. A better question is who did the thinking? The same AI action can help or replace you depending on when it happens and what mental work it performs.

The amount of AI use is not the main issue. A long conversation can still be assistance if you are testing your own argument. A single copied paragraph can be outsourcing if it replaces your thinking.

A four-question decision test

Use this test on one AI exchange at a time. It takes less than a minute.

flowchart TD
    A["📝 Did I make my own attempt first?"] -->|No| X["⚠️ Outsourcing risk"]
    A -->|Yes| B["🧠 Can I explain the key idea without AI text?"]
    B -->|No| X
    B -->|Yes| C["🔍 Did AI act as critic or option generator?"]
    C -->|No, it supplied the thinking| X
    C -->|Yes| D["🖋️ Can I verify and rewrite the suggestion in my voice?"]
    D -->|No| X
    D -->|Yes| E["✅ Helpful assistance"]

The four questions:

1. Did I make my own attempt first?
2. Can I explain the key idea without AI text in front of me?
3. Did AI act as a critic, questioner, or option generator rather than a substitute author?
4. Can I verify the suggestion and rewrite it in my own voice?

If you answer no to question 1 or 2, treat the use as outsourcing until you fix it. If you answer no to question 3 or 4, slow down and do the missing thinking. One no is a warning. Two or more no answers usually mean the tool has taken over the task. The test is a pause button, not a purity score.

Practice: run the test on three capstone moves

For each move, decide whether it is assistance or outsourcing.

A. You write a rough thesis, then ask AI: "What assumptions does this thesis rely on?" You use two questions to revise the thesis yourself. B. You have no thesis, so you ask AI: "Write a thesis about social media for my essay." You paste the result into your draft. C. You write a paragraph, ask AI to fix grammar, then read every change and reject the ones that alter your meaning.

Click to reveal a model answer

A is assistance: you made an attempt first, AI acted as a critic, and you kept the judgment. B is outsourcing: AI supplied the claim before you did the thinking, and the paste removes your ownership. C is assistance if you review each change and keep your meaning. Grammar help becomes outsourcing when you accept edits you cannot explain or when the tool rewrites your ideas, not just your sentences.

Gray zones: paraphrase, brainstorming, and structure

Most real decisions sit between clean assistance and obvious outsourcing.

Paraphrasing. If you wrote the idea and ask AI to make it clearer, you are probably fine, provided you check that the new version still says what you mean. If AI invents the idea and you polish the wording, you are outsourcing. The test is simple: can you explain and defend every sentence as your own?

Brainstorming. Asking AI for possible counterarguments can be assistance if you already made your own list and want to find gaps. It becomes outsourcing when the AI list becomes your argument and you never decide which points are strongest.

Structure. AI can critique your outline: "Where is my evidence thin?" That is assistance. AI becomes the author when it creates the outline and you fill in blanks without deciding the order of reasoning yourself.

A useful habit is to keep AI in a role. A critic, questioner, option generator, or error checker keeps you in charge. A ghostwriter, substitute thinker, or answer machine takes the work away from you.

Here is the quick version: helpful assistance usually comes _____ your own attempt, not before it. The core question is not whether you used AI, but _____. If you cannot _____ an AI-supplied idea without the tool, treat it as an outsourcing risk. AI works well as a _____ and poorly as a _____.

Practice: rewrite one prompt two ways

Choose one gray zone from your capstone: paraphrasing, brainstorming, or structure. Write two prompts. One should invite assistance. One should invite outsourcing. Then mark which prompt you would actually use and why.

Click to reveal guidance

A strong assistance prompt gives AI your material and asks for specific feedback, for example: "Here is my outline. Which claim lacks evidence, and what question should I answer next?" An outsourcing prompt asks AI to produce the missing thinking, for example: "Write an outline for my essay on social media." Your answer should name the difference in sequence and cognitive work, not just in tone.

Keep a short usage trail

For each major AI exchange, note four things: what you asked, what you used, how you verified it, and what you changed. This is not a confession. It is evidence that you remained the author. A three-line version is enough:

• Asked AI to challenge my thesis about school phone bans.
• Kept one counterargument, rejected two that misread the evidence.
• Checked the policy report myself and rewrote the paragraph in my own words.

This habit also prepares you for the capstone reflection, where you will explain your process, not just your final product.

Review deck: assistance or outsourcing?

Front: What one question captures the line between helpful assistance and outsourcing? Back: Who did the thinking?

Front: What is the first question in the decision test? Back: Did I make my own attempt first?

Front: Which comes first in helpful assistance, your attempt or AI feedback? Back: Your own attempt comes first, then AI feedback.

Front: What does cognitive work mean in this lesson? Back: The mental effort of generating, selecting, connecting, and judging ideas.

Front: What role keeps AI on the assistance side? Back: Critic, questioner, option generator, or error checker.

Front: What role turns AI into outsourcing? Back: Ghostwriter, substitute thinker, or answer machine.

Front: Why is a single copied paragraph a risk even if you used little AI? Back: It can replace your thinking and remove your ownership of the argument.

Front: When is AI grammar help assistance? Back: When the meaning is already yours and you review each change.

Front: When does AI brainstorming become outsourcing? Back: When its list becomes your argument and you never judge the points yourself.

Front: When is AI outline help assistance? Back: When it critiques your outline and you keep control of the reasoning.

Front: What does a usage note record? Back: What you asked, what you used, how you verified it, and what you changed.

Front: Does using AI at all mean you are outsourcing? Back: No. The issue is whether AI replaces your thinking and authorship.

Guardrails for the final stretch

• Draft before you prompt. Even five messy sentences give AI something to react to and give you a record of your own thinking.
• Ask for questions, not answers. "What would a skeptic challenge?" keeps you in the author's seat.
• Use the two-no rule. If two test questions fail, stop and rewrite before you use the output.
• Do not treat rewriting as ownership. Changing a few words is not the same as understanding the idea.
• If you would be uncomfortable explaining the exchange in your reflection, it probably crossed the line.

What to carry into the capstone

• AI use is not automatically outsourcing. Sequence, cognitive work, and ownership decide.
• The four-question test is a pause button, not a purity score.
• Gray zones become clearer when you name AI's role and keep a short usage trail.

In the next lesson, Capstone: Draft, Critique, Revise, Reflect, you will apply this test to your own draft, critique, revision, and final reflection.