Hacker News new | past | comments | ask | show | jobs | submit login
Goodnight. Sleep Clean (nytimes.com)
174 points by petethomas on Jan 12, 2014 | hide | past | favorite | 59 comments

> One day, scientists might be able to successfully mimic the expansion of the interstitial space that does the mental janitorial work so that we can achieve maximally efficient round-the-clock brain trash pickup.

We can ask the same question the article poses initially about why we still sleep. Why hasn't this evolved yet? What are the disadvantages of this 'always-GCing' approach? In a best-case scenario, it might just be some sort of historical limitation that no longer applies* (just as our tendency to get fat is a result of an approach to resource conservation that made sense in the ancestral environment, where food was scarce and opportunities to feed intermittent, but does not in modern civilization, where food is nearly always available). In a less appealing scenario there might be some tradeoff that causes us to think less efficiently when this interstitial space expands.

[*] The extra fluids have to come from somewhere, which would mean they're not available for other tasks. Similarly, digestion is expensive--blood flows to the intestines after a meal, and if this is to happen then (since in the short term, you have a fixed amount of blood) it's less available for fueling your muscles. (And if you force yourself to be active right after a meal there's contention for these finite resources. Ever go for a run right after eating?)

Though perhaps there's less fluid involved here than there is with the intestines.

Evolution is actually quite horrible at engineering. Since it can only hill-climb, we end up with ridiculous designs stuck in local maxima. Evolution made video cameras out of jelly, a breathing tube that must be crammed full of solid foods every few hours, and (as Neil deGrasse Tyson put so delicately) a sewage plant next to an amusement park. No organism ever evolved radio communication, or the ability to transmit nerve signals faster than 0.000001c.

Still, why 8 hours of sleep instead of 2? The best explanation I've heard is famine resistance. Sleep conserves energy; until recently, humans were calorie-limited. Also, there's probably an advantage to remaining quiet and still in the dark, when human eyesight is at its worst and predators roam.

Forcing a requirement to think about providing a safe place to sleep results in at least some thinking about safe places to be awake in, and think about defenseless youth in general. If you have to think about finding a safe place to sleep every day, occasionally thinking about a safe place to give birth shouldn't be as challenging. Sounds convenient for long life and successful reproduction.

Also you can tell who hasn't had kids... without kids sleeping (and sleeping more than the parents) the older adults and teens in the tribe would go bonkers and be unable to accomplish anything. Best thing that ever happened to a newborn was sleeping alot so mom can get some food in her, etc. I guess as you get older you could ramp it down and maybe teens could simply stop sleeping, this is probably a secondary effect at most.

> there's probably an advantage to remaining quiet and still in the dark, when human eyesight is at its worst and predators roam.

That's a bit too anthropocentric view. Also other animals sleep, even nocturnal animals.

Worms sleep too - and they don't have brains that need cleaning.

Wikipedia suggests that also in nematodes, "sleep is necessary for changes in the neural system".


Why take human engineering out of 'evolution'? Biological evolution developed our brain and manipulative appendages, which were then used to make better video, comms, etc. Humans, and everything we do, are still 'naturally evolved', by definition.

"How many legs does a dog have, if you call the tail a leg? Four. Calling a tail a leg doesn't make it a leg." — Abraham Lincoln

Side note: according to some sleuthing by the snopes crowd[1], this quote predates Lincoln's presidency by several decades, with the first recorded usage being somewhere around 1834. He may have used the anecdote himself, but it seems to have been a popular saying in the abolitionist movement in general.

[1] http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_t...

Mostly because... I have no idea why. I mean we've discovered something very close to spears being used by people since ... well forever in terms of modern history.

And hunting being very important to the needs of certain homo sapiens sapiens populations would very much agree with your assessment. One advantage that we hold over primates is our ability to throw. Combine that with making spears and our ancestors quickly becoming apex predators makes sense.

What about robustness, adaptability, and learning?

Those are qualities of natural life that I find to be extremely well-engineered, and sadly lacking from most (all?) man-made engineering solutions.

Your brain has 200 billion-ish neurons with hundreds of trillions of connections between them. Right now, reading these words, your visual cortex is doing edge detection and motion filtering. That requires FFTs and other complex computations.

Quick! What's 392 * 7374?

99.9% of people on the planet need pen and paper to figure out the answer. Our oh-so-adaptable brains can't seem to make a few neurons available for simple math. Heck, our brains can't store more than 7-9 items in working memory.

We're mostly unaware of our limitations because everyone else is similarly hobbled. But when we figure out a way to improve human minds, it will be a bigger game changer than anything that has come before.

I'm sorry, I don't follow your point.

I was responding to "Evolution is actually quite horrible at engineering. Since it can only hill-climb, we end up with ridiculous designs stuck in local maxima". In my opinion, several qualities of natural life compare favorably with man-made engineering (which is the only other type of engineering I can think of for the purpose of comparison), therefore I disagree that "evolution is actually quite horrible at engineering", at least without some kind of qualification.

I don't understand the purpose of your 1st or 4th paragraphs; I understand the 2nd and 3rd as saying something like "human brains are limited", but don't see how it connects to my response.

Could you clarify for me? Sorry that I didn't follow and thanks.

The reason why people can't do that calculation in their heads is because of our reliance on computers and calculators.


In fact you can pull up articles with a quick search about how people can attain pretty good mental math capabilities through training. It's hard work, but it's proof of the adaptability of the human mind.

I don't know I already thought the computer was the first "lever" for the mind that we've had. The computer has been pretty game changing for society, has it not?

I learned multiplication long before computers were commonplace, and when four-banger calculators were still quite expensive. The reason I can't do that calculation in my head (well, I could but it would be faster to grab a pen) is that I don't have enough stable registers. I'd guess maybe two registers that won't get corrupted during the calculation. They're not big registers, either.

Regardless, I'm not saying that there isn't a technique that would allow me to improve. I'm saying my inability has nothing to do with ubiquitous electronics.

> The reason why people can't do that calculation in their heads is because of our reliance on computers and calculators.

No, we were never well equipped to do such computations, we didn't suddenly get worse at it with the invention of calculators. Consciously aware mental calculation is a very recent thing on an evolutionary time scale. I think even 10-20k years ago no one did any mental calculations beyond the most trivial arithmetic operations.

OTOH we have a pretty cool native physics engine in our brains and can solve pretty challenging problems such as predicting trajectories and modeling fluid flow. These skills were much more useful in our evolutionary past than arithmetic or algebra.

still nature's creations are 1000s of times better than any created by man. what is the best of what we created? iphone, pc, car?

> (just as our tendency to get fat ...

I suggest that the tendency to get fat is protecting us as a species on an evolutionary scale.

It's unwise to believe we could ditch that just because we haven't needed it for the last 2 generations or so: evolution cares little about a 2 generations scale.

Nice job anthropomorphising evolution. Unfortunately that path leads to ruin.

Evolution has no foresight. Evolution can’t predict the future. Evolution is firmly rooted in the here and now. Doing something now (or not doing something now) because of a prediction of the future is something only humans can do, not evolution. Evolution knows no future.

Evolution knows no future, but it still makes sense to talk about evolution being prepared for a possible future. In particular, what ronaldx is saying is quite close to the accepted explanation for sexual reproduction. Sex produces, and maintains, a great deal of variety within the population that you won't see in an asexually reproducing one. It also imposes a very large penalty to the speed at which the population can grow (males, roughly half of any sexual population, cannot reproduce at all). It is felt that, historically, sexual populations have been so much more able to weather changing circumstances, because of the variability that their lifestyle ensures, that it must have outweighed the susbstantial reproductive penalty -- since sexual organisms have taken over virtually every ecological niche in the world. And this is for an adaptation the entire purpose of which is to guard the population against the future.

In summary: variability in the gene pool is a form of future protection, and one that has been specifically selected for. Be careful before you shout "fallacy!"

What I claim is that this iteration of humans have survived catastrophic black swan events in the past, and that catastrophic black swan events will happen in the future (in a very-long-term, i.e. evolutionary, timeframe).

Living humans don't have any reasonable concept of that, but evolution does have a memory of that[0].

I trust evolution to cover us for this type of future event better than I trust human prediction.

Being rooted in the here and now is something I would ascribe to humans before evolution: humans typically think as individuals on a time scale of 100 years - whereas if something has not been needed for 100 years, this does not illustrate that it is no longer of evolutionary importance.

[0] You can claim I am anthropomorphising here (but not in the earlier comment), but this is sometimes called an analogy, and is an accepted way of using the English language to describe concepts in a clear and simple way. Please ask any geneticist to translate this into evolutionary terminology if you remain unsure.

> "evolution cares little about a 2 generations scale."

I'd take that a step further and say that evolution doesn't care at all!

The tendency to get fat isn't protecting us, it's just something that happened to be an attribute (randomly) of those of our ancestor who didn't die.

Sure, it happened to be an attribute (randomly) of our ancestor who reproduced before dying (not randomly).

> […] why we still sleep. Why hasn't this evolved yet?

A number of animals can put one half of the brain to sleep while the other half remains alert¹.

As for eliminating sleep or shortening it; for such “trait” to evolve there would (beyond being physically possible and the proper mutations having occurred) need to be an advantage that would cause more individuals with the trait to survive (and pass on their genes).

Considering that we have been sleeping in trees and caves, and in groups, there’s probably not a significant larger chance of someone requiring less sleep to make it through the night.

¹ http://en.wikipedia.org/wiki/Unihemispheric_slow-wave_sleep

alternatively, the genes that allow for this would just need to be the dominant ones, the recessive ones that we have now being just the result of a lot of bad luck.

I could well be completely coincidential. It may not affect the rate of procreation at all.

Another possibility is that sleeping conserves resources, because we don't burn as many calories while asleep.

Philosopher Nick Bostrom makes the same argument more generally and detailed here: www.nickbostrom.com/evolution.pdf

We shouldn't expect easy congitive enhancement through drugs unless there is a strong reason it wouldn't have been advantageous in the ancestral environment that doesn't matter now (e.g. much higher calorie usage).

I don't have the will to read a 40 page pdf at the moment, but my initial reaction to that argument is that evolution is not perfect and does not produce the same results in all species. Some animals hibernate for months, while some sleep much less time than we do. Not to mention that it seems like a cop out that avoids any possibility of progress through technology.

Why should we expect an easy cure to syphilis in a drug if evolution has not found one? I don't know, but penicillin still works pretty darn well.

In fact, you shouldn't have expected an easy cure to syphilis. It's a dramatic exception. The vast, vast majority of problems the body has are not easily solved by a simple chemical. This is born out by the tremendous difficulty of finding effective drugs which don't carry unacceptable side effects.

I suspect it's tied to metabolic limitations. It's likely difficult for the brain to provide enough energy to support higher order functions and also do this cleanup efficiently at the same time.

My theory is it has to do with the eyes. The optical input capacity of a human per day has just got to be an incredible stress on the brain.

Then blind people wouldn't need nearly as much sleep, and I don't believe that to be the case.

Consider that blind people often have their other senses enhanced by the body to compensate for the lack of sight. Perhaps it balances out.

I think there has been a study where it was found that the senses aren't actually enhanced, but they just pay more attention to them.

Optic nerves carry thousand to millions of times more information than either audial or touch (e.g. Compare AAC to H264 per second).

Even much enhanced, other senses don't provide even remotely as much information.

Maybe it depends on what type of blind people?

I found this oddly phrased:

"the sleep researchers I spoke with agree that there’s no evidence that aided sleep is as effective as natural sleep."

Is there evidence that it isn't as effective as natural sleep? Why not say there's no evidence either way?

It blows my mind (pun intended?) that we still don't really know why we need sleep, even though we sleep for like 20% of our lives.

We know some stuff, like if you don't sleep, bad stuff happens to your body and you don't form memories effectively, but like the article says we haven't figured out why it's so essential.

20% (4.8 hrs)? I don't know about you, but for me it's closer to 33%.

Haha ya, I guess that's really on the low end. I was thinking of old people who seem to sleep like 4 hours a night for the last 20 years of their lives. Plus people are generally sleep deprived.

Old people pretty much are sleep deprived, though. Their systems are falling apart (that's aging). I was reading a study once on, I think, melatonin in the elderly, and it showed a graph of the 24-hour melatonin secretion cycle for a young healthy sample and then for like age 60+ people; in the young people, the cycle looks exactly as one would expect, with a fall in mid-day, increase in evening, and high levels while asleep during the night and slowly declining to the day levels - in the elderly, it was a flat line. Horrifying.

I'm mostly glad that this will help with Alzhimer's(sp?) research. Being able to work 24hours matters less to me, I rather like to sleep, than treating the terrible memory loss that comes with the disease.

>> "I rather like to sleep"

Personally I agree. I'd also find it interesting to see how removing sleep from our lives would effect things like depression or even how it would just effect people having a bad day. For example if I'm having a terrible day I can't wait to get into bed, fall asleep and end it. The separation sleep provides between one day and another gives the feeling that tomorrow is a fresh start and the problems of today might be less. I'd guess I would be miserable for longer without sleep.

This is by far the most interesting science news I've seen for many months. I am glad it made the HN front page. (And because of the very non-descriptive title, I almost didn't click on it.)

I remember watching a series of videos a few years back by a guy who claimed sleep was for waste disposal and that dreams were because different parts of our brain were shut down at different times. I can't remember his name or where I saw it. Does anyone else remember it or have a link?

Instead of medication assisted sleep, I'll stick to a well organized environment and rhythms.

Real sleep deprivation is really 'funny', after having an allergic reaction to something in my home I couldn't sleep in it for 3 days, it's like being drunk with fever, you lose track of time, balance and get emotionally confused every 30 seconds.

ps: I wish they'd mention 20 minute long naps and how they can feel so "repleting" even though I don't believe a lot can be cleaned through the glympathic system.

I'd be interested to see what effect (if any) modafinil has on this cranial garbage collection routine. Given subjective reports that modafinil allows one to go without sleep without accruing sleep debt for some period of time, maybe this would shed some light either on how modafinil works or what other functions sleep has (or both!).

Curious whether meditation might induce a state where the brain also cleans itself, just like during sleep...

Afaik, that's how they explain why the Elves in some version of D&D fantasy roleplaying only need 4h of sleep/meditation :)

But seriously, I've never noticed anything like that in my own practice. Granted I don't go for stretches longer than about 45 minutes, and I do it every morning when the cleaning has just happened. Or perhaps you'd need a specific additional technique apart from just sitting focusing on breathing. Visualisation techniques can sometimes dramatically change the effect of a meditation session (sometimes, but not consistently).

Additionally, I suggest trying to meditate when you're mildly sleep deprived. You'll find it harder to do, with yourself dozing off or blinking/blacking out occasionally. And while they're both, in some sense, mental states of "no thought", they are also worlds apart: the meditative state is a state of high focus, while the dozing/blinked state is pretty much the opposite.

What I've found in the few occasions I've experienced this (I don't like being sleep deprived so I try to avoid it :) ), is that you come out of such a blink-out state with a tiny bit of a jolt. Now, say I'd apply the same technique as one would use for basic meditation practice: when your thoughts wander away, try to observe that this has happened, congratulate yourself on noticing (instead of seeing it as failure) and gently bring your thoughts back to the focus.

Now I guess (and I'd have to try it out, but I don't want to because it involved sleep-depriving myself) one of two things will happen: One, over time (weeks/months), this process will train myself not to fall asleep, setting up circuits in my brain to keep me from dozing off (hopefully only when I don't want to). Of course the cleaning thing won't happen because you're awake and focused all the time, so you just got more efficient at sleep depriving yourself. OR number two, perhaps more likely, you'll doze off anyway, at some point. I can totally see that happen, dozing off into a half-dream where I dream I bring my mind back into focus and have this amazing meditation, suddenly get weird, in my dreams, and wake up a few hours later. It's the typical kind of trick my brain would pull on me.

Also, I don't quite see meditation like this. True, after 30 mins of meditating I am usually more relaxed and focused than I was before (but not always), but the main thing about meditation for me is just like sports and exercise: it's not so much about the immediate results, but it's training. When I meditate I do it because I am training my mind to keep focus and not drift off to everywhere except here & now, to have that state at least once a day, but to also take that habit onto the rest of the day when I'm not sitting. What you're describing sounds more like meditation as a kind of "pill" to give a direct effect, I find it's not very reliable in that sense, I use it more like training one's back muscles to give more support during the day.

Article title is somewhat misleading, I thought they meant sleep clean as in without going into an intoxicated state, or to actually sleep clean and that this was somehow relevant.

"Sleep cleans your brain" or "Sleep garbage collection" or something. Still probably not good.

The first paragraph of this contains my latest pet peeve - any kind of evo explanation which relates a human behavior or biological process to the hominid era of evolutionary adaptation ... when that behavior is common to all mammals or even all vertebrates.

Apart from "rebooting" body and mind during sleep, I also think that dreams are extremely important as real life's sandboxing; used to evaluate the efficacy of future actions and prepare for them.

... the future's gonna be weird ...

That brain graphic is like a motion illusion when I read down the text; it continually appears to shrink. Did anyone else see that too? It only works when you aren't focusing on the graphic.

Interesting article.

Off-topic: the new nytimes.com website feels a bit buggy on my iPad. None of the top banner button works (of the article page). Are they using JQuery-Mobile or similar slow js framework?

jQuery v1.7.1 as far as I can tell


Let's all just sleep properly, till someone finds a pill which will rejuvenate us in lesser time, and we fool humans overuse/misuse it and crib about technology later :)

why would staying quiet and motionless during nocturnal predator hunting time be an evolutionary disadvantage?

Because you are also completely defenceless. If a predator stumbles across you'll probably already be in pieces by the time you're aware what's happening. Ideally you would want to stay quiet and motionless but also awake and alert.

Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact