Sleep on it: How the brain processes many experiences, even when 'offline'

Cool to see that this worked well for someone. Super hard to force the key insight in a problem to magically appear given more time sunk into it. Big weakness of mine honestly, and requires a lot of self-awareness to pull myself out of a problem-solving rut. I like the idea of hacking sleep - do you find yourself priming your mind with the problem before nodding off? Curious how a bedtime wind-down routine factors into how effective this is.

While programming Java full-time, I found myself waking in the middle of the night with some big chunk of my brain grinding away on coding issues. Not good.

I do the same with the NYT crossword puzzles. The puzzle for a day is released at 10 pm NY time (except the Sunday and Monday puzzles are released 4 hours earlier) the day before, which is 7 pm in my time zone.

I start the puzzle at the end of my day. If I get stuck and cannot finish it before I fall asleep, probably 95% of the time when I take a look again sometime the next day I immediately see answers to several of the things I was stuck on and finishing the puzzle goes smoothly.

This pretty much describes my method of working as well. Often at the end of the day, when I’m not feeling like I have enough energy to solve anything new, I’ll spend some time just looking over some new problems I’ll have to solve in the next few days. I don’t try to solve them. I’m just “loading them into my head”. The next morning I usually find some potential solutions waiting for me. Thanks, brain! I’ll repeat that process a few times until I have them solved. Unfortunately, while this works great in real life, it doesn’t work for interviews.

I don't think it needs to be as "structural" as you're describing here. Simply "work on something else if you're stuck, come back to it after a day, two days, or longer" has long been one of my secrets to get stuff done.

I once worked somewhere that allowed you to work only on one ticket/task, before you were allowed to move on to another. Completely dumb policy for so many reasons.

I think it was Thomas Edison who said, “Never go to sleep without a request to your subconscious.”

Excellent advice.

I don't even need the coffee - I wake up having solved the problem.

Generally around 2 or 3 in the morning, just when I need it the most.

I think just about everyone has a similar anecdote.

I was stuck on particularly tricky part of a musical piece I was learning - hours and hours of practice and I just couldn't get it down. Went to bed, woke up and was able to nail it within the hour.

Sleep is underrated.

This sounds like me but exactly at the opposite hour - late at night

That being said the pre req being that it's a problem I've already known about (and likely slept on)

I had a college professor suggest this and I 100% agree. I think of it as loading the problem in your brain. Then sleep on it and you will make a lot more progress in the morning than if you had just spent the same total time in one sitting.

I have the same feeling during coding at night.

In the end, a good sleep is what make it different!

a lot of people ditch the coffee for psilocybin

I don’t think it can be compared to caffeine in its efficacy at all with the mushrooms being far more useful, but do note that both caffeine and psilocybin rely on anecdotes for cognitive performance, despite one being able to be studied as a non controlled substance

This reminds me of what I learned about myself during my years spent at the university. I observed that in the morning my brain is better at understanding new concepts. Mornings were the best time for me to practice and improve problem solving, but I tend to remember less details of what I come across. However, at about 2pm my brain appears to switch to memorizing mode, where I struggle with problem solving compared to the morning, but I will remember a lot more of what I read. I structured my learning activity leveraging this observation. Even to this day (am 46) I can feel the same tendency, e.g., if a problem seems somewhat difficult, I just wait until the next morning, if I can, only to find it easy to come up with some solution that seemed out of reach the previous evening. Also, I try to do most of my reading at night (well, life with a family doesn't leave a whole lot of options for timing anyway).

This resembles the oldest self-optimisation-trick I've ever been exposed to: Soak in a topic short before going to bed, avoid distractions before actually falling alseep, and "watch" your brain learn overnight. I think I was aroun 10 when I was told this trick. And to this date, its about the only tangible memory optimisation technique I've ever put to use. Nothing ever came close to its effectiveness.

I have a similar hack but my breakthroughs typically happen on my morning run (after your mentioned good night of sleep). No ear buds, etc. Just a casual pace and giving my mind time and space to work.

This works. It also works fine for programming tasks. If you feel your struggling or making a slow progress on your current task, switch to a different task.

When you come back to the first one your subconscious has usually processed it quite a bit.

Even during an exam, having a micro-nap or going for a walk is beneficial

This is amazing to hear.

Have you studied the hippocampus memory of hummingbirds much?

Their glucose is consumed to preserve their extraordinary 3d spatial memory for all their food sources.

---

https://pubmed.ncbi.nlm.nih.gov/22357941/

"...analyses reveal that the HF in hummingbirds is significantly larger, relative to telencephalic volume, than any bird examined to date..."

-- IMO the ability of the Hippocampal spatial 'compuiting' - is what consumes the glucose in what is their 'GeoPU' - and so to be able to think in a 3D volumetric space - not vector space - allows for such precise control where the positional coordinates of a food source have a heavy weaight in 3D and Temporal Memory - which is the same as how it can navigate in a 3D point space with its hovering...

It Computes to Live and it Lives to Compute.

GPU folks should be studying hummingbirds.

Especially for AI patchfinding with sensor awareness - as some hummingbirds migrate ~2,000 miles from Chile to Canada.

https://www.hummingbirdsplus.org/nature-blog-network/ruby-th...

Now note how frequently this critter needs aerial refuling, and it needs to know the best and most efficient path to hop all the food-check-points over 2,000 miles

https://i.imgur.com/3MvzmX9.png

Hummingbirds can fuel expensive hovering flight completely with either exogenous glucose or fructose

https://besjournals.onlinelibrary.wiley.com/doi/full/10.1111...

>By foraging frequently and fuelling hovering flight directly with ingested monosaccharides hummingbirds avoid the energetic tax associated with the cost of synthesis of fats from these sugars prior to their oxidation. Remarkably, hovering hummingbirds are able to utilize fructose and glucose equally, a physiological feat which no mammals are thought to match, and one that suggests novel physiological capacities for the oxidation of fructose by active muscle tissues in hummingbirds. The data presented here indicate hummingbirds enhance net energy intake though specialization of diet, behaviour, and, uniquely, metabolic physiology.

-- now imagine the ripples that are running through the hippocampus thats maintaining this level of efficient precision of a Body that has near instant acceleration and precisice altitude control in a 3d volume.

Hummingbirds are the most amazing critters.

We should be studying humming birds glucose control through the HF, not rats.

What do you mean by agent? Human, brain, some component of the brain, etc.?

I am also a habitual walker, though I don't typically get eight miles in. I often listen to books and podcasts while walking and at times wonder if I'm doing myself a disservice by not just letting my mind wander. One the one hand, it is about the only time I allow myself a solid hour to listen to something, but on the other maybe it's time better spent giving 'space to work' as you say.

I'm amazed by people who can simply just..decide to sleep with some crisis hanging over their head.

I really wish it was a skill I had, because I'm well aware of how badly I perform when I haven't had enough sleep.

I inevitably end up with the worst of both worlds - trying to sleep at an early hour, can't sleep due to the gears in my brain ticking over, getting frustrated at the fact I can't sleep, which makes it even harder to fall asleep.

I end up just as tired as I would have been if I had simply stayed up for 4 more hours, but at least then I could have actually done something in that time besides stew over the fact that I can't sleep.

Do you have more details on your strategy?

Your process sounds like something a writer might say. Are you a writer?

I am saying based on how writers/poets etc are depicted. Waiting for ideas to arrive, sometimes they keep they keep waiting for it, sometimes it hits them on a random event.

Rick Rubin has a similar philosophy, he describes it in his book “The Creative Act”. Has he been an influence on you?

How can you fit your method together with the many nuances around making a living? Surely you must structure your day to day somewhat, no?

this is my approach as well :peace:

Funnily enough I wonder if a burst of adrenaline after learning is effective. Kind of tricking your brain into thinking it just survived a dangerous situation and must retain whatever lead up to it.

Would love to see some research on it, but this reminds me of the structure of a traditional yoga asana practice and I've wondered the same thing about rest after a learning/training session. Typically at the end of a practice session of yoga asana (and sometimes in between postures) you'll take savasana and just rest in open awareness. A teacher of mine would always say this is the most important part of class, to let everything integrate in.

When I started training Muay Thai I found that I often would "lose" a lot of what we had worked on in class, and then started taking savasana right when I got home. In that not-quite-sleep state my brain would replay what we worked on.

Sometimes if I'm really bashing into a wall with some coding problem I'll just take savasana for 10-15 minutes and get back to it. I feel like it helps.

The whole setup of the modern school day is dumb. It's too early for many growing minds. And it doesn't balance what's taught with a chance to reflect on it. And the school year should be the whole year with more breaks rather than a summer time off.

It would be pretty cool to have some sort of siesta for students in the day. Maybe one of those private schools would do it.

This was somewhat the thesis of my dissertation.

I suggested cognitive fatigue was an adaptive construct that biases people to go offline and replay their memories, and that this was decision-theory optimal from a learning / reward perspective.

https://mayank-agrawal.com/papers/AgrawalMattarCohenDaw21.pd...

When I feel fatigue coming on, I just lie on a sofa with some soundscape playing on my noise cancellation headphones.

I close my eyes and allow my mind to drift. Similar to the phase before falling asleep. Most of the time, this mind drifting is enough but sometimes I fall asleep - both states help.

I do something similar. I ask "does this feel like there is a solution space, is this probably solvable?" If there is I pursue it or ruminate on it (mainly incubate it). If not, I ask why not or reframe it or move it to "wish list" (my step below "roadmap").

Creativity benefits from incubation, setting it aside to see what reframings and new ideas pop up.

and about remembering stuff. When I want to remember something, I read it some time before sleeping. After reading I remember nothing, but after waking up it magically appears in my memory. Very useful

I think writing part from cache to hard disk is part of NREM sleep cycle (from the book Why we sleep)

REM is responsible for dreams, where it matches experiences from cache to long term storage

IIRC

> REM sleep is inhibited when you use sleeping pills or alcohol.

In world war 2, some commanders found that a shot of brandy after a difficult action could help stave off "combat fatigue". It would be interesting if their solution actually helped by reducing traumatic memories forming.

WW2 should be a poster child for getting enough sleep. There were multiple mistakes made by both sides due to leaders not having slept for days. Admiral Kurita might have pressed his advantage in the Battle off Samar, or Admiral Halsey might have avoided the 2 typhoons his task groups encountered. Of course, it's easy for me to sit here and type that without experiencing the overwhelming pressure of war.

I love this! When dealing with a difficult problem, I like to think about it in my head as falling asleep. I find I don’t necessarily have the solution upon waking, but I will understand the space better, and be less fixated on whatever my initial thought was.

Some go as far as believing that plants have consciousness. Which isn’t surprising considering slime moulds:

> Slime molds have a variety of behaviors otherwise seen in animals with brains.[0]

[0] https://en.wikipedia.org/wiki/Slime_mold

Would like to hear your Architecture more

You just described my ideal situation.

Covid shut down was the only time I've been fully remote, and I was MASSIVELY productive. Because I could manage my own schedule around how my brain works.

This is how I do it. 10-2, twice a day. Sleep 7.5 hours a night. Works great for me.

Are we sure there is a difference?

Research has shown that we form memories related to things that happen around us in our sleep and are conscious in some ways. We don't form episodic memories of sleep time and it's likely that what it's like to be conscious while asleep is quite different from our waking experience even aside from dreams.

> Are we unconscious when we sleep

I don't think that is the right question to ask. Sleeping person is unconscious almost by definition. They don't respond to stimuli and they don't do much of anything besides breathing and just laying there.

> do we merely lose the capacity to remember?

That is a better question. Some people remembers their dreams. Sometimes people can remember stimuli which happened while they were asleep. (For example it happened multiple times to me that noises happening around me where incorporated into my dreams in various forms.) So if sleeping were just amnesia then these dream memories would need to be explained somehow.

I think there's a medical difference between unconscious and sleeping. At least I don't think being forced unconscious by something (a bonk on the head) is like sleeping

I guess it depends on how you define consciousness. I think by most definitions, we definitely are unconscious when we are asleep.

Ever entered a room and immediately forgot why you came in? Dream might be much the same.

Memory could be attached to context. Change the context extremely and you lose the associated memories.

And dream is an extreme change of context.

well, "was selected-for over trillions of iterations", they said, "... by algorithms which were themselves selected-for"

"randomly" makes it sounds like some kind of miracle!

I’m guessing you are saying it is hard to believe that evolution creates incrementally greater complexity over billions of years?

Reminder that our best estimates of spacetime curvature Omega puts it very close to 1.0: the universe is very close to flat, so if the universe is indeed not spatially infinite (possible) it is at least 100s of times larger in radius than the observable universe, which contains about 10^21 stars. [1]

If the universe is spatially infinite, then all possible quantum states exist in infinite copies. So yeah, “it happened randomly” is not absurd. In fact, I find it a bit strange that any educated person would think otherwise.

[1]: https://en.wikipedia.org/wiki/Shape_of_the_universe

If it was designed by a creator then who or what created that creator? Something has to begin out of nothing!