Hi, this is Ray.

Let me describe a moment that, looking back, I think every learner has experienced and almost no one has interpreted correctly.

You're working on something hard. A problem you can't solve. A concept that won't click. A piece of writing you can't get unstuck on. You've been at it for two hours, and you're going nowhere. The harder you push, the worse it gets. Eventually, exhausted, you give up and walk away. You go take a shower. You go for a walk. You stare at the ceiling. You let your dog out and watch them sniff a single blade of grass for 11 minutes like it contains the secrets of the universe.

And then, with no warning, somewhere in the middle of that completely unproductive activity, the answer arrives. Fully formed. Out of nowhere. As if it had been waiting in another room for you to stop banging on the door and finally just leave so it could come find you.

I have had this experience hundreds of times. So have you. So has every person who has ever wrestled with a hard problem. And we have, almost universally, drawn the wrong lesson from it. The lesson we tend to draw is "huh, weird, glad that worked out, now back to the grind." The lesson we should be drawing is: that thing that just happened? That isn't an accident. That isn't a quirk of consciousness. That's a fundamental mode of how learning and problem-solving actually work, and once you understand it, you can deliberately use it as a tool instead of stumbling into it once a year.

Today's newsletter is about that. About why relaxation isn't the absence of learning. About why, sometimes, walking away from the problem is the work. And about how to actually structure your study sessions to take advantage of one of the most underused features of the human brain.

AI will write 90% of code by the end of 2026, and only 10% of developers will stay relevant.

The engineers in that 10% aren't smarter. They just know how to use AI.

We put together the exact playbook Anthropic engineers use to make sure you're in the top 10%.

Sign up for The Code and get access to:

Join 250K+ engineers

Let's start with the technical name for what I just described. In cognitive psychology, it's called the incubation effect, and it's been one of the most replicated findings in problem-solving research for over 100 years.

The basic phenomenon: when you're stuck on a problem, taking a break and engaging in something else (even something completely unrelated) often allows you to come back to the problem and solve it more easily than if you'd just kept grinding. As one research summary put it, the incubation effect refers to the reliable boost in problem-solving performance that occurs after a break or distraction period, compared with continuous concentrated effort, and from Archimedes' bathtub Eureka to modern hackathons, taking time away often unlocks insight that eluded conscious focus. This isn't folklore. It's been measured, replicated, and studied across decades. Scientists have given participants problems, had some of them work continuously and others take breaks, and consistently found that the break-takers solve more problems, or solve them faster, or both.

The interesting question is WHY. There have been two main theories, and the truth seems to be a combination of both.

The first theory is unconscious processing. Your brain keeps working on the problem in the background even when you're not consciously focused on it. As the same summary noted, cognitive processing continues covertly; associative networks keep recombining until a viable solution surfaces. When you walk away from the problem, your conscious mind disengages, but networks of related neurons keep firing, exploring connections, trying combinations. The default mode network (which I've covered in a previous newsletter) seems to be a major player here. Your brain doesn't actually stop working on the problem when you stop working on it. It just shifts to a different mode of working.

The second theory is forgetting-fixation. When you're stuck on a problem, you're often stuck because your mind is locked into a wrong approach. You can't see new possibilities because old ones are blocking the view. As one major review of incubation research summarized, creative responses can be blocked by more dominant ones, a fixation effect, and forgetting dominant responses can enable incubation effects, and forgetting fixating responses can occur via temporal delays, retrieval inhibition, and context shifts. Translation: the break lets you forget the wrong approach, which lets the right approach finally show up. You're not "incubating" so much as you're letting the bad solution fade so the good solution has room to emerge.

The honest scientific answer is probably "both, depending on the problem." Some problems benefit from the unconscious processing. Some benefit from the forgetting. Either way, the practical implication is the same: when you're stuck, walking away is often more productive than pushing harder. The break IS the work.

It's not just about solving stuck problems. There's a separate body of research showing that even short, deliberate rest breaks during ordinary studying improve learning outcomes.

A 2023 study published on rest breaks and learning tested this directly. Researchers had students complete attentionally-demanding arithmetic tests (designed to deplete cognitive resources), then either proceed straight to a math lesson or take a 5-minute rest break first. According to the researchers, when students learn or solve problems, attentional resources are depleted, and rest breaks may restore cognitive functioning in support of learning. Students who took the break learned the subsequent material better. Five minutes. That's all it took to change the cognitive state enough to improve learning outcomes.

This connects to a broader principle from attention research. Your directed attention (the kind you use to focus on study material) is a finite resource that gets depleted with use. Push it too long without a break and your performance declines, your encoding gets sloppy, and your retention suffers. The break isn't lost time. It's restoration time. The math is closer to "investing 5 minutes to make the next 30 minutes 50% more effective" than to "wasting 5 minutes you could have studied through."

A separate study examining incubation specifically in educational contexts found similar effects. The researchers, examining students playing an educational physics game, found that students who failed to solve a problem on first attempt and then took a break were more likely to solve it on a second attempt, demonstrating clear incubation effects in computer-based learning environments. Not just for ancient Greek bathtub revelations. For modern students learning physics. The effect is real, replicable, and applies to actual learning material, not just artificial lab puzzles.

Here's where it gets nuanced. Not all breaks are equal. The research has spent a lot of effort figuring out what TYPE of break produces incubation effects, and there are some specific findings worth knowing.

Mind-wandering breaks tend to outperform demanding breaks. A break where your mind is free to drift (a walk without a podcast, a quiet shower, light housework, staring out a window) tends to produce better incubation effects than a break filled with another demanding task. The reason: the demanding task occupies the cognitive resources that would otherwise be working on your problem in the background. Scrolling Twitter is technically a break from studying. It's a terrible incubation break because it floods your attention with new stimuli. Real relaxation > productive relaxation > scrolling.

Length matters, but not as much as you'd think. Studies have found incubation effects with breaks as short as 5 minutes and as long as several hours or overnight. Short breaks help with smaller problems and immediate fatigue. Longer breaks (especially overnight, including sleep) help with bigger creative problems. Match the break length to the difficulty of what you're stuck on.

The break needs to actually disengage you. Just sitting at your desk staring at the same problem isn't a break, even if you're not actively working on it. Real incubation requires a context shift. Get up. Go somewhere else. Do something different. The physical and cognitive separation matters.

Sleep is the king of incubation. Multiple studies have shown that sleeping on a problem produces some of the largest incubation effects of any kind of break. As one summary put it, sleep, particularly the REM stage, is crucial for aiding cognitive functions like memory consolidation and creative thinking, and during sleep the brain organizes and processes information from the day, helping with memory recall and comprehension. If you're really stuck on something, the answer might genuinely be "sleep on it." Not as a metaphor. As a strategy.

Here's what I actually do, after years of trial and error and many forehead-meets-desk moments. Steal what's useful.

The "deliberate walk-away" technique. When I notice I've been stuck on something for more than 20-30 minutes without progress, I stop. I don't push harder. I don't tell myself "just five more minutes." I close the laptop, get up, and go do something else for at least 15 minutes. Walk. Make tea. Pet the dog. Stare at a wall. The rule is: no problem-related thinking, no productive activity. Just a real shift. When I come back, I solve the problem maybe 60-70% of the time, often within minutes. The other 30%, I'm at least making forward progress instead of grinding in circles. The walk-away has become a tool, not a defeat.

Scheduled relaxation breaks during long sessions. Every 50-90 minutes of focused work, I take a real 10-15 minute break. Not "check email" break. Real break. Walk around. Look out the window. Make food. The Pomodoro people figured this out in the 80s and they were right. The session quality after the break is consistently higher than what you'd get from grinding through. I track this. The math always favors the break.

Deliberately leaving problems unfinished. When I'm working on something complex (a piece of writing, a hard concept I'm trying to understand, a problem I'm thinking through), I sometimes deliberately stop in the middle, even when I'm not stuck. I leave the problem on the cognitive stove to simmer overnight. The next morning, I almost always come back with new ideas, new connections, or a clearer sense of what to do next. This is incubation as a routine practice, not just an emergency tool.

The "shower notebook" rule. Because so many ideas come during shower-walking-staring-at-ceiling moments, I keep a way to capture them everywhere. Note app on my phone. Notebook by the bed. Voice memo while walking. The ideas come whether you're ready for them or not. Most of mine evaporate within minutes if I don't catch them. Catch them.

Sleep-on-it for big stuck problems. For anything I've been wrestling with for more than a day, I genuinely use sleep as a tool. I review the problem briefly before bed (not in a stressed way, just a "here's what I'm chewing on" way) and then I let it go. The next morning's first thought, surprisingly often, contains some part of the answer. This isn't magic. It's REM sleep doing what REM sleep does. I just stopped fighting it.

I want to be clear about something, because the "relax to learn" framing can get distorted into "do nothing and you'll learn by osmosis." That isn't the point. The incubation effect requires that you've ALREADY DONE THE WORK. You can't incubate a problem you haven't engaged with. The break is productive specifically because of what came before it.

As one of the early studies on incubation noted, the break's only function is to divert the solver's attention from the problem, thus releasing her mind from the grip of a false organizing assumption, and this enables the solver to apply a new organizing assumption to the problem's components upon returning to the problem. Note the word "returning." The break only works because you go BACK to the problem afterward. Without the return, there's no incubation. There's just relaxation. Both have value, but only the structured version produces the learning effect.

So this isn't "relax instead of studying." It's "relax AS PART OF studying." The cycle is: focus, get stuck or saturated, break, return, repeat. Skip any step and the system stops working.

Here's the meta-point. We've inherited a deeply flawed model of what learning looks like. The model says learning is uninterrupted focus, willpower, and grinding through. The model treats breaks as moral failures. The model treats relaxation as the absence of work, something you earn after enough productivity.

The actual science says something quite different. Learning is a cyclical process that requires ALTERNATING phases of focused engagement and disengaged rest. The disengaged phase isn't optional. It's where the consolidation happens, where the connections form, where the insights emerge, where the resources replenish for the next focused phase. Trying to learn without rest periods is like trying to build muscle by lifting weights 24 hours a day. Not just less effective. Actively counterproductive. The recovery IS the building.

If you've been treating your relaxation as guilty pleasure or wasted time, I'd gently suggest reframing it. Your downtime isn't time off from learning. It's the second half of the learning. The shower thoughts, the walking ideas, the "huh, I just realized" moments in the middle of the night… those aren't bonuses. They're outputs of a system that's been running quietly while you weren't looking.

So when you sit down to study tomorrow, build the breaks in. Schedule the walk-away. Plan the staring-out-the-window. Trust that when you come back to your work, your brain will have been working on it the whole time. The relaxation was the work. You just couldn't see it happening.

Even Frodo had to stop walking sometimes. The hobbits had second breakfast for a reason. The best ideas in Middle-earth probably happened during pipe breaks at the inn, not during forced marches. We could learn from this.

Keep learning (and keep relaxing),

Ray