AI as Friction for Reflection Support in Ideation
This paper argues that generative AI in design ideation should be reframed from a tool for removing friction to a "friction agent" that actively supports reflection-in-action, thereby helping designers build and communicate rationale rather than simply accelerating output.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine you are in a giant, buzzing workshop where people are trying to invent new things. This is the world of design ideation, a field where scientists and artists study how humans come up with fresh ideas. Usually, we think the best way to be creative is to move as fast as possible, like a race car driver who never hits the brakes. We assume that if we remove all the bumps and delays, we will get more and better ideas. But there's a hidden part of the process called reflection, which is like a quiet moment where you stop and ask, "Why did I just do that?" and "Does this actually make sense?" Without these pauses, you might end up with a pile of cool-looking toys, but you won't know how they work or why you built them. This paper asks a big question: What if we stopped trying to make AI (Artificial Intelligence) the fastest race car, and instead made it a speed bump that forces us to think?
The authors of this paper, Janin Koch, Xiaohan Liao, and Géry Casiez, argue that most AI tools for creatives are designed to do exactly what we expect: they remove all the friction. They generate images or ideas instantly, let you delete bad ones with a single click, and keep the workflow smooth and fast. The authors suggest this is actually a trap. They propose that AI should act as "friction"—not a frustrating wall, but a helpful pause. Think of it like a bouncer at a club who doesn't let you in until you tell them your name and why you're there. By making you stop and explain your choices, the AI helps you build a rationale, which is the story behind your idea. This story is crucial because it helps you remember why you made a choice and allows you to explain it to your friends or teammates later.
The paper suggests that when AI makes everything too easy, designers might end up with hundreds of ideas but no memory of why they picked any of them. It's like having a magic wand that conjures a thousand cakes, but you don't know which one tastes good or why you wanted cake in the first place. The authors point out that in group projects, this is a disaster. If you can't explain the "why" behind your sketch, your team can't build on it, argue with it, or combine it with their ideas. The paper suggests that AI should introduce "structured pauses." For example, before an AI deletes an idea you don't like, it could ask, "Why are you throwing this away?" or before it shows you a new picture, it might ask, "What are you trying to solve right now?"
This isn't about making the work harder or slower in a bad way; it's about making the thinking deeper. The authors suggest that this "friction" helps designers practice reflection-in-action, which is a fancy way of saying "thinking while you are doing." Instead of just churning out results, the AI becomes a partner that nudges you to build the reasoning behind your work. The paper doesn't claim to have solved this problem yet or built a perfect tool. Instead, it offers a new way of looking at the problem and asks the community of researchers to figure out exactly how much "friction" is good, how to make it feel natural, and how to make sure it helps groups work together better. It's a suggestion to slow down the machine so the human mind can catch up and build something truly meaningful.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.