My dream journal solved a problem I'd been stuck on

A few weeks ago I was stuck on something at work. I'd been trying to figure out why a certain user segment was churning out of the app at a point that didn't match any of the obvious explanations. The data wasn't wrong, it just wasn't making sense, and I'd been staring at the same retention chart for three days.

Then I had a dream about swimming. I was underwater in a pool, holding my breath, and every time I came up for air the pool had gotten deeper. I wrote it down in my journal the way I always do. The pool, the breathing, the depth changing. Didn't think much of it until I stepped away from the screen later that morning and the connection hit me: the users dropping off were the ones who'd never set up their first journal entry — they never got in. The depth in the dream was depth of engagement. My brain had been working on it without me.

I'm not claiming the dream literally solved my research problem. But something about writing it down and then moving my body shook loose a thought I couldn't get to while hunched over my laptop. That was enough to make me curious about whether this kind of thing is a real phenomenon or just a story I was telling myself.

It's real. And the science is more interesting than I expected.

Edison was onto something

Thomas Edison used to nap in a chair while holding a metal ball in each hand. When he dozed off and his muscles relaxed, the ball would drop, clang against the floor, and wake him up. He'd grab whatever idea had been floating through his mind in that in-between state and write it down.

For a long time this was just a good anecdote. Then in 2021, Delphine Oudiette at the Paris Brain Institute tested it for real. She gave 103 participants math problems that had a hidden shortcut baked in. Most didn't find it. Then she had them recline for 20 minutes, holding a drinking glass the way Edison held his ball. The ones who drifted into N1 sleep, that first, lightest stage where your thoughts start to loosen, were nearly three times more likely to find the hidden rule than the ones who stayed awake. And nearly six times more likely than the ones who fell into deeper sleep.

There's a specific state, right at the edge of consciousness, where your brain makes connections it can't make when you're fully alert. Oudiette said she was "quite surprised that almost no scientists had studied this period" until recently. That makes two of us.

Researchers are now engineering dreams

This is where it gets strange. In February 2026, neuroscientists at Northwestern published a study where they cued people's dreams during REM sleep. Karen Konkoly and Ken Paller gave 20 participants a set of difficult puzzles, each one paired with a specific piece of music. Most went unsolved. Then they went to sleep in the lab, and the researchers played the puzzle-linked music during REM.

Seventy-five percent of participants dreamed about the cued puzzles. And those puzzles were solved at a 42% rate the next day, compared to 17% for the ones that weren't cued. The dreams were doing something.

At MIT, Adam Haar Horowitz and his team built a device called Dormio that does something similar at sleep onset. It's a glove with biosensors that detects when you're entering N1 and prompts you to dream about a specific topic. In one study, they prompted people to dream about trees. Afterward, those participants wrote stories that were 43% more creative than people who napped without prompts. And 78% more creative than people who just stayed awake.

This is the part that unsettles me a little. Engineering someone's dream content feels like it should be further off in the future than it is.

You don't need a lab to do this

In 2022, Raphael Vallat and colleagues at the Lyon Neuroscience Research Center published a study comparing people who frequently remember their dreams to people who rarely do. On a standard creativity test, where you list unusual uses for ordinary objects, the high recallers averaged 8.2 ideas versus 7.2 for low recallers. The gap got wider when they looked only at the most original responses. The high recallers also had more cross-talk in the brain's default mode network, the part that's active when your mind wanders.

There's also an older study where participants spent 27 days writing down their dreams each morning. A control group wrote about waking events instead. Both groups improved on raw creativity scores like idea volume, but when researchers scored for qualitative stuff like emotional expressiveness, humor, and fantasy, only the dream journalers improved. The researchers' theory was that going back and forth between dream thinking and waking thinking every morning loosened something up creatively.

That second study is the one that stuck with me. No gloves, no lab, no music piped in during REM. Just people writing down their dreams every morning for a month, and ending up thinking differently because of it.

What I've been doing differently

Since the swimming pool dream, I've been trying something. When I wake up with a dream that has any connection to something I'm working on, even a loose, metaphorical one, I mark it in my journal with a small asterisk. Not to analyze it in the moment, but to flag it for later.

In about six weeks I've flagged nine dreams. Most of them didn't lead anywhere. Two did. One was a dream about trying to tune a radio where the stations kept shifting, which I wrote about on a morning when I'd been stuck on a logic problem in the app. Reading the entry back a few days later, I noticed I'd described the problem differently in dream language than I had in my notes, and the dream framing helped me see what I'd been overlooking.

The other was a dream about my grandmother's kitchen where she was sorting things by color but the colors kept changing. I still don't know what that one was about. But I noticed while writing it down that I'd been grouping the data by the wrong dimension. Not because the dream told me that directly. Because the act of writing loosened something.

What I still don't know

I don't know if I'm just noticing connections because I'm looking for them now. Confirmation bias is a real thing, and I'm a sample size of one. The lab studies are convincing in controlled settings, but my morning journal is not a controlled setting. It's me, half-awake, with my phone, trying to hold onto a pool that keeps getting deeper.

What I do know is that my dream journal has become something slightly different from what it was six months ago. It used to be a record. Now it feels more like a workspace. A place where my sleeping brain leaves notes for my waking one, and sometimes those notes are useful.

I'm going to keep flagging dreams and see what the pattern looks like at three months. If you journal your dreams and you've noticed something similar, I'd be curious to hear about it.