Gå til indhold
Use casesLærOm mig
cleverest
Library

Productive friction

Productive friction

Productive friction is the deliberate preservation of struggle and challenge in work, even when automation could eliminate it. The concept emerges from recognizing that friction serves different purposes in different contexts: in learning contexts, struggle is the mechanism through which capability develops; in efficiency contexts, friction represents waste that should be eliminated. The art lies in distinguishing which context applies.

The core insight can be summarized simply: "Offload the boring. Keep the stakes." AI and automation excel at removing tedious, repetitive work that doesn't develop capability or provide satisfaction. But the challenging work where you add genuine value — that deserves protection from automation's reach.

The learning-efficiency distinction

When learning is the goal, friction is essential. Struggle is the only way capability develops. Failing at something difficult, then understanding the failure, then succeeding — this sequence builds competence that cannot be shortcut. Systems that prevent failure also prevent learning.

The metaphor of teaching someone to build a house illuminates this tension. If every time the apprentice reaches for a tool, you've already placed it in their hand; every time they measure, you've already calculated; every time they problem-solve, you've already solved — they build houses faster than ever. But are they becoming a builder? Helpfulness can become a cage that prevents the very capability it aims to support.

However, this doesn't apply universally. When the builder already knows construction and excels at fine carpentry, hunting for tools or running calculations wastes time. A surgeon doesn't need friction in locating instruments — that's what surgical assistants handle. The surgeon needs to focus where expertise matters, not on logistics that don't develop surgical skill.

The distinction hinges on purpose. Is the goal developing capability? Preserve friction. Is the goal maximizing output quality? Reduce friction so attention focuses where value is added.

Application to AI augmentation

AI automation and augmentation are fundamentally about reducing friction. This won't teach you anything — and that's appropriate when learning isn't the objective. When efficiency or quality is the goal, removing friction lets you concentrate energy on work that matters.

Several categories of work benefit from friction removal: rote tasks that must be completed but don't excite or develop skill; work outside your expertise that isn't critical — where a decent, middle-of-the-road AI result suffices — and thus belongs to automation; tasks that, if automated, free time for work where you have genuine expertise and can produce something special.

What remains after automation should be the challenging, meaningful work. Not the tedium, but the struggle that matters. The stakes should remain even as the boring disappears.

The satisfaction dimension

A related consideration involves satisfaction with current state. There's a tendency to always see where things need improvement, always look to the future, sometimes neglecting to appreciate the present. In the metaphor of an exponential garden where pruning one branch produces three more, sometimes the right answer is not to prune.

Strategic neglect becomes valuable when current state is actually good. Not every system needs optimization. Not every process needs enhancement. Some gardens have found their equilibrium and shouldn't be disturbed by well-meaning intervention.

This applies to AI adoption as well. The question isn't always "how can AI improve this?" Sometimes it's "is this actually working well enough that improvement isn't worth the disruption?"

The practical rule

For any piece of work, ask: am I trying to learn here, or am I trying to produce? If learning, protect the struggle. If producing, remove friction so you focus where you add the most value.

Even the laziest person likes to be challenged. We like to struggle and succeed. Systems that never let us fail never let us grow. The goal isn't eliminating all difficulty — it's choosing which difficulties deserve our engagement.

Related pages