B5 · Generative AI and Conversational Assistants

AI as a Thinking Partner

Using AI as a thinking partner means using it to question, test and extend your own thinking, instead of using it to produce answers so that you do not have to think. The first builds your abilities; the second lets them fade.

6 min Beginner Jump to the activity

What you need: Nothing. This page starts from the beginning, with no programming and no formulas.

Definition

A calculator made nobody worse at mathematics, or did it? What will a tool that writes and reasons do to our minds?

Using AI as a thinking partner means using it to question, test and extend your own thinking, instead of using it to produce answers so that you do not have to think. The first builds your abilities; the second lets them fade.

In simple words

You can use a bicycle or a taxi to cross town. Both get you there. The bicycle makes your legs stronger every time. The taxi does not, and if you take it for every journey, your legs grow weak. AI can be either. Asking it to challenge your idea is the bicycle. Asking it to have the idea for you is the taxi. Taxis are useful. A life with no cycling at all is not healthy.

Why it matters

Thinking is a skill that grows with use and shrinks without it. A tool that can think for you is the first technology that offers to do the whole job. How you use it will shape how capable you are in ten years.

Key terms

Thinking partner
Someone, or something, that helps you think better by questioning and responding to your ideas.
Cognitive offloading
Handing a mental task to a tool, such as using a phone to remember numbers.
Critical thinking
Examining ideas carefully, weighing evidence and reasons before accepting them.
Devil's advocate
Someone who argues against an idea on purpose, to test how strong it is.
Socratic questioning
Teaching by asking questions that lead the learner to work out the answer.
Over-reliance
Depending on a tool so much that you cannot manage without it.

Explained in detail

Handing work to tools is not new

People have always passed mental work to tools. Writing let us stop memorising long stories. The printed calendar let us stop tracking dates in our heads. Mobile phones took over the telephone numbers that people used to know by heart. Psychologists call this cognitive offloading. It is often sensible: it frees attention for more important things. But there is a cost. What you stop practising, you lose. Few people today can recite many phone numbers, and many drivers who always follow navigation never learn the layout of their own city.

What is different about generative AI

Earlier tools took over narrow tasks: remembering, calculating, finding the way. A generative tool can take over the central ones: understanding a text, forming an argument, deciding what to say. These are the abilities that education exists to build, and they underlie everything else. Early studies of people using AI assistants for writing and problem-solving point in a consistent direction. Those who let the tool do the work finish faster and remember less of what they produced. Those who use it to check and challenge their own work tend to keep the benefit.

Ways to make it a partner

TechniqueWhat you askWhat it does for your thinking
Explain it back"Here is my explanation of how a vaccine works. What did I get wrong?"Shows the gaps in your understanding
Devil's advocate"I think school uniforms are a good idea. Give me the three strongest arguments against."Makes you meet the other side
Question me"Do not give me the answer. Ask me questions that help me work it out."You do the reasoning; it steers
Find the holes"Here is my plan for the science fair. What could go wrong?"Tests your plan before reality does
Widen the view"What have I not considered?"Adds options you can then judge yourself

Think first, then ask

The order matters more than anything else. If you ask the tool before you have thought, its answer becomes your starting point, and it is very hard to think past it. Your ideas bend towards whatever it said. If you form your own view first, even a rough one, you have something to compare with. You can notice where the tool disagrees and decide who is right. A useful rule is to write down your own answer, however short, before you open the tool.

Keep your own voice

A language model produces what is typical. If millions of people ask it for an essay on friendship, they receive millions of similar essays. Writing and ideas shaped heavily by AI drift towards a smooth, sensible average. What you can add is exactly what the average lacks: your own experience, an unexpected comparison, an honest opinion, a detail from your street or your family. Protect those. They are what make work worth reading.

Skills to keep without the tool

  • Reading a long text and working out what it says.
  • Writing a clear paragraph from an empty page.
  • Doing arithmetic and estimating whether a number is reasonable.
  • Remembering the core facts of the subjects you study.
  • Judging whether an argument holds together.

These are the skills that let you check what a tool gives you. Without them you cannot tell a good answer from a bad one, and you become dependent on a machine you are unable to judge.

How it works

  1. Think first. Write down your own answer or idea, however rough.
  2. Ask the tool to respond to your thinking, not to replace it.
  3. Compare. Where does it differ from you? Who has the better reason?
  4. Check any facts it adds.
  5. Decide for yourself, and say it in your own words.

Try it yourself

Thinking with AI, not through it

Meera must decide what her group should build for the science fair.

Thinking partner, or thinking replacement?

Who does the thinking in each case?

  • "Summarise the chapter so I need not read it."

  • "I solved it this way. Is there an error in my steps?"

  • "Ask me questions until I can explain this myself."

  • "Here is my argument. What would someone who disagrees say?"

  • "Just tell me which option to choose."

  • "Write my opinion on this topic."

Common mistakes

  • Asking the tool before forming any view of your own.
  • Accepting its suggestions without weighing them.
  • Letting basic skills go because the tool can do them.

Check yourself

Question 1

1. What does it mean to use AI as a thinking partner?

Question 2

2. Why should you think before you ask the tool?

Question 3

3. What is cognitive offloading?

Summary

  • AI can strengthen your thinking or replace it, depending on how you use it.
  • Think first, then ask it to challenge, question and check your ideas.
  • Keep the basic skills that let you judge what it gives you.

Finished reading?

Sources: UNESCO: Guidance for generative AI in education and research · AI4K12: Five Big Ideas in AI