Productive friction

Garden · not yet reviewed Emerging: new and still settling

Deliberately adding a small amount of effort to working with AI, such as drafting first, predicting before the reveal or asking it to argue back, so you think the problem through before accepting an answer.

Also called: desirable difficulty · think first · study mode · learning mode · guided learning · AI that challenges

Connections

Understand first

Ideas this one builds on.

• Cognitive offloading (garden)

Related

Often used together or easily confused.

• Human in the loop (garden)
• Centaurs and cyborgs (garden)
• Clarifying questions (garden)
• Sycophancy (garden)

Skills that use it

• practice-coach (garden)
• /brainstorm (garden)

Blueprints that practise it

• Ask AI to find where your draft misleads (garden)
• Ask AI to play devil's advocate on a team decision (garden)

Resources that teach it

• How AI assistance impacts the formation of coding skills (garden)
• The AI Fluency Index (garden)
• Anthropic Education Report: How university students use Claude (garden)
• Generative AI without guardrails can harm learning (garden)
• Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning (garden)
• How to Use Claude Code as a Second Brain (garden)
• Your Brain on ChatGPT (garden)
• A Conversation with Dr. Vivienne Ming, author of Robot-Proof (garden)
• A top researcher says a new divide is emerging in AI use, and most people are on the losing side (garden)
• Vivienne Ming on hybrid collective intelligence, building cyborgs, meta-uncertainty, and the unknown infinite (garden)
• In Conversation... with Vivienne Ming (garden)
• Robot-Proof: When Machines Have All the Answers, Build Better People (garden)
• The AI That Asks Better Questions: Dr. Vivienne Ming at SIM 2025 (garden)
• 15 Times to use AI, and 5 Not to (garden)
• Against "Brain Damage" (garden)
• Choosing to Stay Human (garden)
• Wharton Generative AI Labs Prompt Library (garden)
• Instructors as Innovators: A future-focused approach to new AI learning opportunities, with prompts (garden)
• AI Should Challenge, Not Obey (garden)

Competency

• Use responsibly
• Check the result

Find productive friction on the commons map ↗

You may know it from

• learning science desirable difficulties Spacing, self-testing and generating answers feel harder but make learning stick, as long as the learner can succeed.
• project management a pre-mortem Imagine the project failed and list why, before anyone commits.
• fitness a personal trainer adding weight
• cognitive science hybrid intelligence (Vivienne Ming) Ming's test for any tool: does it leave you more capable once you switch it off? She argues for AI that questions you instead of handing over answers.

In products

The same idea, under different names in different tools.

Provider	What they call it
OpenAI	ChatGPT study mode
Anthropic	Claude learning mode
Google	Gemini Guided Learning
Wharton Gail	Devil's Advocate prompt (Wharton Generative AI Labs) (CC BY 4.0; a critic that asks one question at a time.)

In plain words

Most AI tools are designed to remove effort. Productive friction adds a little back, on purpose, where it pays off. A few examples: write your own rough answer before asking. Predict what the AI will say before you read it. Ask it to argue against your plan. Ask it to explain its reasoning, then restate it in your own words.

Getting the amount right matters. With too little, you accept polished answers without thinking; with too much, you stop using the tool. Good friction is light by default, heavier for high-stakes or unfamiliar work, and easy to skip with “just do it” when you’re short on time.

Several providers now offer learning or study modes that ask questions instead of handing over answers. They help most when you want to learn the subject yourself.

Example

Before asking AI to prioritise her team’s projects, a director ranks them herself in two minutes. The AI’s ranking differs on one project. Talking through why, she realises she was underweighting a funder deadline, and adjusts. She trusts the final ranking more because it’s hers.

How it connects

• The design answer to cognitive offloading.
• The antidote to sycophancy: AI that challenges instead of confirms.
• Clarifying questions and human-in-the-loop approvals are light forms of it.

Try it

Next time you ask AI a question you could partly answer, write your guess first, including how confident you are, and only then read the AI’s answer. Do this three times this week and notice when your guess beat the AI.

---

Read this card as plain Markdown: /concepts/con-productive-friction.md. Know a parallel from your own field, or a product that implements this? Add it.