Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Good sleep, good learning, good life

supermemo.guru

Post reply on HN
Print view View original

71–80 of 125 posts

Re: Good sleep, good learning, good life

#71

Post by m463 »

muzani wrote: post #58

This has been one of my favourite articles for years, on the reading list, but never had the time to fully read such a behemoth. Some stuff I got out of this: 1. The ideal sleep pattern is sleeping when tired and walking up without an alarm. 2. Biphasic is recommended. Homeostatic is how easy it is to initiate sleep, circadian is how easy it is to maintain it. Circadian cycle is usually nightly, homeostatic peaks wit…

naps are sometimes great, but sometimes if I don't set an alarm I end up feeling lousy. Don't know if it means sleep or something else is out of whack and I need a nap too much.

the pandemic has let me play with sleep and sometimes it's to let it gradually drift forwards an hour each day and wrap around. (schedule meetings are the only trouble spot to that experiment)

Re: Good sleep, good learning, good life

#72

Post by allenu »

wenc wrote: post #67
bd_beats wrote: post #26

Earlier quoted context omitted.

Anki's SRS algotihm is based on SM2 and is woefully outdated in comparison to SM's SRS algo. Additionally SM is much more feature rich once you get into incremental reading and fully optimizing your learning experience. Sure the UI isn't a nice flashy Electron.js app, but SM is a titan compared the terminate that is Anki.

So, SRS algorithms are such that the more you try to get the optimality, the more the returns diminish because there's significant variability in the human population (bias-variance tradeoff sort of applies). The Ebbinghaus curve itself is a very crude approximation. See this: https://languagelearning.stackexchange.com/questions/3757/wh... (I'm a language learner and an active user of SRS) The notion of an outdated a…

I’m in strong agreement with your points. It always surprises me when I read people trying to come up with a more optimal algorithm for testing one’s retention when there are so many factors out of one’s control that would influence the forgetting curve. For instance, the notion of judging your ability to recall something and providing it as input to Anki is such a gross approximation. Choosing the wrong value from 1 to 5 would affect your recall schedule considerably.

In my opinion, what’s more important to a recall strategy is just pure habit. You have to practice regularly, and anything that makes the software easier to use for the long term is something that would have a greater effect than the optimality of the recall algorithm. (Incidentally this is why I’m working on a flash card app that is easier to use, to help ensure one comes back to it regularly.)

Re: Good sleep, good learning, good life

#73

Post by qznc »

briefcomment wrote: post #53
chrisweekly wrote: post #47

Earlier quoted context omitted.

See also "The Happiness Equation" which boils down to "Be happy first." NOT: Hard work -> success -> happiness RATHER: Happiness -> best work -> success

Intuitively, I don’t doubt that, but certainly there are unhappy “successful” people right? My tentative explanation for this is that many people luck themselves into “success” and then fight tooth and nail to keep it.

A different angle I sometimes see is Unhappy -> Ambition -> Success.

However, these types of people might not be happy in the end. For example, I remember Elon Musk saying "you wouldn't want my life."

Re: Good sleep, good learning, good life

#74

Post by qznc »

supersrdjan wrote: post #14
jmnicolas wrote: post #8

Piotr Wozniak is a fascinating guy, if you want to read more about him: https://web.archive.org/web/20150613033756/http://archive.wi... Is someone using his software here?

When you get hooked on SuperMemo and embrace the idea of incremental reading, it becomes the primary software on your computer. You end up spending more time in SM than you do in the browser. People use it not just to memorize flashcards, but to read articles and books... even to write and edit their own articles and books... track their sleep... plan and execute their daily schedule... keep track of tasks and priori…

Sounds like Emacs. Does it have a Lisp inside? Can it send email yet?

Re: Good sleep, good learning, good life

#75

Post by mrlala »

briefcomment wrote: post #53
chrisweekly wrote: post #47

Earlier quoted context omitted.

See also "The Happiness Equation" which boils down to "Be happy first." NOT: Hard work -> success -> happiness RATHER: Happiness -> best work -> success

Intuitively, I don’t doubt that, but certainly there are unhappy “successful” people right? My tentative explanation for this is that many people luck themselves into “success” and then fight tooth and nail to keep it.

For sure, infact "successful" people are probably in general less happy because they do not prioritize happiness. Sure they might be able to more easily achieve some artificially (vacations, new stuff). But that isn't the same as truly learning how to be happy.

Re: Good sleep, good learning, good life

#76

Post by japhyr »

muzani wrote: post #58

This has been one of my favourite articles for years, on the reading list, but never had the time to fully read such a behemoth. Some stuff I got out of this: 1. The ideal sleep pattern is sleeping when tired and walking up without an alarm. 2. Biphasic is recommended. Homeostatic is how easy it is to initiate sleep, circadian is how easy it is to maintain it. Circadian cycle is usually nightly, homeostatic peaks wit…

I committed to living without an alarm clock in college, 20+ years ago. I fought long battles with my snooze button, and realized I was spending hours each morning hitting the snooze button. I realized if I just didn't set an alarm I'd probably get up later than I usually set my alarm for, but earlier than I typically got out of bed. It worked; I started waking up refreshed, and happy to get out of bed in the morning.

When I had an important exam or something else I really didn't want to miss, I'd set an alarm for a while. But then I just went all in, and never set an alarm. That made me pay more attention to when I was going to bed at night, and the quality of my sleep. When I started working I would set an alarm for about half an hour after I typically woke up, just in case, so I wouldn't be late for work. But after six months of that, I stopped setting my backup alarm as well. I was late for work maybe twice in five years, and it was never significantly late, and it was at times where my body really needed sleep and it probably helped me avoid getting sick. These days I only set an alarm if I have to catch a 6am flight, or if I'm getting up super early to climb a mountain or something.

If you're battling your alarm every day, I highly encourage you to experiment with not setting an alarm.

Re: Good sleep, good learning, good life

#77

Post by sradman »

joveian wrote: post #35
sradman wrote: post #23

Earlier quoted context omitted.

Regarding 1: Sleep loss lethality is caused by gut ROS in mice and flies [1] > ...researchers demonstrate in flies and mice that sleep loss leads to an increase of reactive oxygen species (ROS) in the gut, and that neutralization of ROS can prevent oxidative stress and prolong the lifespan of sleep-deprived flies. Various biological systems seem to have recovery mechanisms associated with the sleep cycles. These cycl…

Thanks for the link. Lots of companies have products with claimed effects or measurment of sleep but personally I suspect they are almost always complete BS. I've used a recording pulse oximeter when sleeping a number of times and the pulse seems to reliably predict when I get to sleep but nothing else, not even when I am awake vs. sleeping. IMO, there is no possible way a pulse measurement could replace an EEG. I am…

I forgot to comment on the following important point:

> the pulse seems to reliably predict when I get to sleep but nothing else, not even when I am awake vs. sleeping

My experience with the Zeo made me realize that predicting sleep vs wake is hard to do. The modern multi-sensor devices tend to use the accelerometer derived movement together with historical sleep times to predict this. The early Fitbit devices tracked sleep with the motion sensor alone. In the middle of my normal sleep cycles, if I listen to a podcast or read from a smartphone/eReader while remaining mostly motionless, both the Zeo EEG and the Garmin PPG derived EEG will interpret that time as REM sleep.

I wonder if the professional sleep labs can accurately detect the difference? The true test will be when sensors/algorithms in casual activity trackers can detect when we nod off or nap during normal waking hours.

Re: Good sleep, good learning, good life

#78

Post by rajlego »

wenc wrote: post #67
bd_beats wrote: post #26

Earlier quoted context omitted.

Anki's SRS algotihm is based on SM2 and is woefully outdated in comparison to SM's SRS algo. Additionally SM is much more feature rich once you get into incremental reading and fully optimizing your learning experience. Sure the UI isn't a nice flashy Electron.js app, but SM is a titan compared the terminate that is Anki.

So, SRS algorithms are such that the more you try to get the optimality, the more the returns diminish because there's significant variability in the human population (bias-variance tradeoff sort of applies). The Ebbinghaus curve itself is a very crude approximation. See this: https://languagelearning.stackexchange.com/questions/3757/wh... (I'm a language learner and an active user of SRS) The notion of an outdated a…

Have you tried SuperMemo? Intervals matter because they correlate with rep loads. The more accurate the intervals, the lower daily rep loads can be.

The SuperMemo creator has spent 3 decades and collected plenty of data, it's hard to imagine that he wouldn't be able to improve on an algorithm from 1990.

Re: Good sleep, good learning, good life

#79

Post by rajlego »

qznc wrote: post #74
supersrdjan wrote: post #14

Earlier quoted context omitted.

When you get hooked on SuperMemo and embrace the idea of incremental reading, it becomes the primary software on your computer. You end up spending more time in SM than you do in the browser. People use it not just to memorize flashcards, but to read articles and books... even to write and edit their own articles and books... track their sleep... plan and execute their daily schedule... keep track of tasks and priori…

Sounds like Emacs. Does it have a Lisp inside? Can it send email yet?

Funnily enough, you can use it to process email (which I have tried, and used): http://super-memory.com/help/e-mail.htm

Re: Good sleep, good learning, good life

#80

Post by sradman »

joveian wrote: post #69
sradman wrote: post #44

Earlier quoted context omitted.

See Garmin Health Announces Sleep Study Results [1]. I used a Zeo [2] mobile for a few years and I thought I had a good feel for what my sleep cycles tended to look like. I bought a Garmin VivoSmart last Black Friday mostly to try the SpO2 and sleep tracking. I shared your skepticism but, anecdotally, the sleep cycles are as good as what I would get from Zeo. [1] https://www.garmin.com/en-US/blog/health/garmin-health…

Hmm... 70% agreement in an unpublished study funded by the company vs a device that is itself maybe 90% agreement with lab PSG. Better than I expected, but I wonder if it is just due to many sleepers having quite predictable stages. "Figure 4 shows the worst result of the cohort (55 of 55), with an accuracy of 49.9%. Note that this user self-reported having a sleep disorder."

> I wonder if it is just due to many sleepers having quite predictable stages

That is what I wondered. The Garmin reports my deep sleep in ways that don't match predictable patterns and do approximate times of random events like being woken by animal noises or sirens. Its much better than I expected and I can see it getting much better with the incorporation of additional sensor/external data.

Post reply on HN