Is Doing Things the Hard Way Better? The Friction Trade

Doing things the hard way is better only when the difficulty is doing the building, and worse whenever it is not. That is the entire answer, and it dissolves the usual hard-versus-easy argument into a sharper question: is this friction constructing skill, memory, or judgment, or is it just costing you time? Effortful cognition, generating an answer instead of looking it up, retrieving from memory instead of rereading, wrestling with a problem before being handed the solution, reliably produces a stronger mind, while pure busywork friction produces nothing but fatigue. So the rule is selective, not heroic: keep the friction that builds, offload the friction that does not. And the stakes of getting this right rose sharply, because as AI makes the easy path nearly free, the productive struggle is exactly the part most people are about to stop doing, and therefore exactly where a durable advantage now hides.

Why does the hard way ever beat the easy way?

Because for the mind, effort is not a cost paid alongside learning, it is often the mechanism of learning itself. Robert Bjork’s lab named this family of effects desirable difficulties, and the research program is consistent: conditions that make acquisition feel harder and slower, spacing practice out, generating rather than receiving, varying the context, tend to produce more durable and transferable learning than conditions that feel smooth and fast. The fluency of the easy way is partly an illusion; it feels like learning while depositing little.

Two specific mechanisms carry most of the effect. The generation effect: information you produce yourself is remembered better than the same information handed to you, and recent work using pupillometry to study the generation effect traces this to the mental effort generation demands, the effort is the signal that marks the material as worth keeping. And retrieval practice: the evidence collected at Retrieval Practice shows that the struggle of pulling a fact out of memory strengthens it far more than the ease of putting it in again. In both cases the difficulty is not friction around the learning; it is the learning, which is why the smooth shortcut so often leaves you with nothing in your head.

So is all hard work worth it? No, and here is the line

The line is whether the difficulty builds something durable in you. Plenty of friction is pure dead weight, manually copying numbers between spreadsheets, fighting a broken tool, re-deriving something you will never need again, and treating that as virtuous is a mistake in the opposite direction. Effort romanticized for its own sake is just inefficiency wearing a halo.

Kind of friction	What it builds	Verdict
Generating an answer before checking	Memory, understanding, transfer	Keep it: the struggle is the learning
Retrieving from memory vs looking up	Durable recall, fluency	Keep it for things you must own
Wrestling a problem before seeing the solution	Reasoning, problem-solving skill	Keep it: this is where capability forms
Manual, repetitive busywork	Nothing; just fatigue and lost time	Offload it: automate or delegate
Fighting bad tools or broken process	Nothing; resentment	Fix or remove it
Re-doing solved, never-reused work	Nothing	Offload it

The practical test for any task: if I do this the hard way, will I be more capable afterward, or just later? Capability-building friction is an investment; time-costing friction is a tax. Most arguments about “the hard way” go wrong by failing to separate the two and then defending or attacking both at once.

What does the energy framing actually tell us?

That thinking is cheap to run and expensive to feel, which is why we avoid it even when it pays. The brain runs on roughly the power of a dim lightbulb, yet hard thinking is subjectively costly, and the research on whether thinking hard burns calories finds the metabolic difference is real but small, the exhaustion of deep work is largely the felt cost of effortful control, not a big energy bill. So the aversion to the hard way is more psychological than physical: the mind treats cognitive effort as something to minimize, and shortcuts feel good partly because they relieve that felt strain.

This reframes the AI era cleanly. Outsourcing cognition to a machine removes the felt strain, which is exactly why it is so seductive and so quietly costly: the strain you are removing is, in the learning cases above, the thing that was building you. The brain’s striking efficiency means the productive struggle is nearly free to perform and enormously valuable to keep, the asymmetry behind the case for the 20-watt supercomputer. The decision is not about saving energy; it is about whether you let the machine spend the effort that would have made you more capable.

How do you decide what to keep hard?

Keep hard the things you need to own; offload the things you only need done. The sorting rule has two questions. First, is this a capability I want to live in my own head, judgment in my field, the skills my work actually rests on, the knowledge I need available without a lookup? Those get the hard way, deliberately, because the struggle is what installs them in your biological knowledge graph. Second, is this something I merely need produced, with no benefit to me from having struggled? That gets the easy way without guilt.

The concrete pattern for the AI age: do the thinking yourself first, then let the machine assist, never the reverse. Draft before you prompt, attempt the problem before you ask for the answer, form your own view before you read the summary, this preserves the generation and retrieval effects while still capturing the tool’s speed on everything downstream. It is First Brain before Second Brain as a daily discipline: the strong internal model is built by friction you chose to keep, and the tools amplify a mind that did the hard part rather than replacing the part that would have made the mind. Building that model deliberately, deciding which struggles are worth keeping across a whole domain, is the project of Building Your First Brain, free for the first 1,000 readers, and it is the same logic that makes low-compute innovation and a wetware renaissance more than nostalgia.

What are the honest caveats?

Several, because “keep the hard way” is as easy to over-apply as “take the shortcut.” First, friction that builds skill for the learner is often just waste for the expert: a surgeon should not re-derive anatomy, a senior engineer should not hand-roll what a library does correctly, so the same task can be desirable difficulty for a novice and pure tax for a master, and the verdict shifts with where you are. Second, not everything needs to be owned, no one can build deep capability in every domain, so most of life’s friction should be offloaded, and the deliberate hard way is for the few areas that matter to you, not for all of them. Choosing universally hard is its own failure, a recipe for doing little, slowly.

Third, there is a real equity and access dimension: “do it the hard way” is easier to prescribe than to afford, since the time to struggle productively is itself a resource not everyone has, and disability, deadlines, and circumstance legitimately change the calculus. The balanced verdict, then: doing things the hard way is better precisely when the hardness is constructing the skill, memory, or judgment you want to keep, and worse the rest of the time. The skill is not toughness; it is discrimination, knowing which friction to guard and which to delete, and in a world rushing to make everything frictionless, that discrimination is itself the rare and valuable thing.

Key takeaways: is doing things the hard way better?

Sometimes, and the test is precise: the hard way wins when the difficulty is the mechanism building durable skill, memory, or judgment, generating answers, retrieving from memory, struggling before the solution, the desirable-difficulties findings, and loses when the friction is busywork that builds nothing. Thinking is cheap to run and only feels expensive, so the aversion to effort is psychological, which is exactly what makes AI’s frictionless path seductive and quietly costly. Keep the friction that strengthens you, offload the rest, do your own thinking before the machine assists, and remember the verdict shifts with expertise, with how much you need to own a skill, and with what struggle you can actually afford.

Frequently asked questions

Is doing things the hard way actually better?

Only when the hardness builds something durable in you. Effortful cognition, generating answers, retrieving from memory, wrestling a problem before seeing the solution, produces stronger, more transferable learning than the smooth easy path, which often feels like learning while depositing little. But friction that builds nothing, manual busywork, fighting bad tools, redoing solved work, is pure waste and should be offloaded. The skill is telling the two apart: keep difficulty that makes you more capable, delete difficulty that only makes things slower.

Why does struggling to learn something work better than being shown?

Because for the mind, effort is often the mechanism of learning, not a cost beside it. The generation effect shows you remember information you produce yourself better than the same information handed to you, and pupillometry research links this to the mental effort generation demands. Retrieval practice shows that the struggle of pulling a fact from memory strengthens it more than rereading. In both, the difficulty is the learning, which is why frictionless methods feel efficient yet leave little behind.

Does thinking hard burn a lot of energy?

Not much. The brain runs on roughly the power of a dim lightbulb, and research on whether hard thinking burns extra calories finds the metabolic difference real but small. The exhaustion of deep work is mostly the felt cost of effortful mental control rather than a large energy bill. That matters because it means our aversion to hard thinking is largely psychological, which is exactly why outsourcing it to a machine feels good while quietly removing the effort that was building us.

When should you take the shortcut instead?

Whenever the task is something you only need produced, not something you need to own. If struggling through it would not leave you more capable, manual data entry, fighting a broken process, re-deriving something you will never reuse, automate or delegate it without guilt. The same applies once you are already expert: re-doing what you have mastered, or what a reliable tool does correctly, is a tax, not a discipline. Reserve the deliberate hard way for the skills and knowledge that matter to you.

Will AI make the hard way obsolete?

The opposite, for the things you want to own. AI removes the felt strain of thinking, which is seductive precisely because that strain was building your skill and memory in the learning cases. As the easy path becomes nearly free, the people who deliberately keep the productive struggle, doing their own thinking before the machine assists, compound a capability the tool cannot give them. The hard way becomes a choice rather than a default, and choosing it where it counts becomes a real edge.