I notice that after a while, without enough usage, these topics simply evaporate from my memory. I know the simple answer is to keep practicing.
I’m interested in knowing how different people do this. Looking for any tips, hacks, etc.
I don't know why nobody else has suggested this.
If you want to learn, really learn something. Self educate, then teach it to someone else. You'll have to really educate yourself on the subject, but once you do.. Blog post, Twitter, Youtube, Linked-In, or whatever platform you want to use. You'll get corrections if you're wrong about something, and hopefully a load of questions that will make you dig further into the topic. The process of teaching it will reinforce it into your memory as you present it in the way you understand, and the feedback you get will also strengthen your understanding even further. You'll become much less likely to forget it over time.
But what if you're like me, and you don't have a social media presence or following to get any kind of feedback? Surely you have friends with similar interests, or know one of those people who likes soaking up knowledge like a sponge. You could use HN here to post something, or post to Reddit's "Today I Learned".
Inevitably, teaching something is the best way to learn more about that thing, and learning more about it is what will make it stick better.
In other words, if you are a "bad" teacher -- if your students do not understand, do not care, or do not otherwise learn from what you impart -- does the mere act of trying to pass on the information work for retaining/refining the information yourself? And if so, is a blog or forum post or drunken rant at a pub any different than a lecture?
I ask because I have found myself expounding on things in informal contexts and learning more about my own ideas as I go, with the supposed intention of telling and teaching others, but my feedback at times has been "dude you suck at explaining this". I do try to adjust according to feedback, but useful feedback is rare, especially when marred by the emotional reaction of "what do you mean I suck?? you sure you ain't just stupid??" etc.
No one wants to fumble in public like an idiot and waste people’s time, so you prepare for the material as well as possible questions.
There ARE some topics I want to remember, and those I try to put in writing. Both because it helps them stick in my mind for longer and because then I have a reference to go back to if I need a refresher. I use my own blog quite a bit to remember ways to do things that I may have figured out earlier and then forgotten.
But in terms of keeping things completely in memory: I have grown to think that, for a healthy mind, the things worth remembering are the things we remember. It's _good_ to remember things you're regularly using. If you're not regularly using them, you may as well forget the details and refresh your memory later if needed. Of course the major caveat being that this is within reasonable limits and not a symptom of an illness or actual serious memory loss.
It is a balance, and losing my memory in general is one of my worst fears. I guess it is a matter of deciding if these topics are beginning to evaporate sooner than you're comfortable with. Do they disappear after a day? A week? A year? If I start forgetting knowledge too quickly for my standards, that's when I'll get worried. It's just a matter of deciding what "too quickly" is for us.
>Hyperthymesia, or highly superior autobiographical memory (HSAM), is a condition that leads people to be able to remember an abnormally large number of their life experiences in vivid detail.
>Hyperthymestic abilities can have a detrimental effect. The constant, irrepressible stream of memories has caused significant disruption to Price's life. She described her recollection as "non-stop, uncontrollable and totally exhausting" and as "a burden".[1] Price is prone to getting lost in remembering. This can make it difficult to attend to the present or future, as she is permanently living in the past. Others who have hyperthymesia do not display any of these traits, however.
Your brain is not wired to remember things that you don't use. If you want to continue remembering, you have to continue engaging with the topic, and a lot of the advice in this thread is IMO just that (just, with a specific regimen / kind of engagement). It's completely natural to forget.
In fact, I'd go further and suggest embracing the forgetting and relearning cycle. If you really need to revisit the topic, you can relearn it again, and it'll be faster than the first time. And maybe you'll forget another time -- that's OK too, you can pick it up again even faster. In the end it's the discipline of relearning (the particular topic, and in general!) that actually continuously hardens. IMO there are parallels between this mindset and SRS -- though without the shame when you forget or miss a day of practice.
[1] A Mind for Numbers (book): https://barbaraoakley.com/books/a-mind-for-numbers/
[2] Learning How To Learn (Coursera course): https://www.coursera.org/learn/learning-how-to-learn
[3] Make It Stick (book): https://www.retrievalpractice.org/make-it-stick
[4] Augmenting Long-Term Memory (blog post): http://augmentingcognition.com/ltm.html
[5] SuperMemo Guru (website): https://supermemo.guru/wiki/SuperMemo_Guru
[6] Nelson Dellis (YouTube): https://www.youtube.com/channel/UCnINhoHWuVjUDXp7dav5e3A
I second the use of Anki--it's a great tool. Other (non-free) options include SuperMemo and IDoRecall.
Update: After reading the API section of this post, I found this very last part interesting in the context of OP's question[0]. I thought it may be worth highlighting to the OP that even a heavy user of a memory system recognizes that knowledge we don't regularly use may not be worth remembering.
[0]"A more challenging partial failure mode is Ankifying what turn into orphan APIs. That is, I'll use a new API for a project, and Ankify some material from the API. Then the project finishes, and I don't immediately have another project using the same API. I then find my mind won't engage so well with the cards – there's a half-conscious thought of “why am I learning this useless stuff?” I just no longer find the cards as interesting as when I was actively using the API.
This is a difficult situation. I use the rule of thumb that if it seems likely I'm not going to use the API again, I delete the cards when they come up. But if it seems likely I'll use the API in the next year or so, I keep them in the deck. It's not a perfect solution, since I really do slightly disconnect from the cards. But it's the best compromise I've found."
Paying attention to what knowledge I (or top performers) use most often in practical projects / everyday work has helped. I also try to identify fundamentals; i.e. chunks of knowledge that experts in a field have identified as critical to understanding that field.
Anki is a popular app that uses spaced repetition. I personally wasn’t quite satisfied with any of the options out there so I decided to build a new app to solve this problem. Feel free to check it out: https://activerecall.com/ It allows you to create flashcards using a rich text editor with support for things such as code and latex blocks. I’m using it daily both to learn new software development concepts and for general knowledge.
Even with a spaced repetition-tool however there is a lot of nuance that goes into the process of acquiring new knowledge. A couple of pointers here:
1. It’s important to distinguish the learning step and the act of remembering what you’ve learned. Learning for example means using a new programming concept to the point where you understand it. Now that you have that context, you can encode it in the form of question and answer in a spaced repetition based system to ensure that you will remember what you’ve learned.
2. Some things that you want to remember are purely factual, while some require deeper understanding. In the latter case, you’ll want to structure your flashcards in a way that you’re solving a problem when answering it rather than just stating facts. This ensures that you’re actually remembering how to solve something rather than simply stating what the solution is.
3. Some things that seem like they are worth remembering when you first put them into your system will later turn out to be time sinks. Don’t be afraid to just get rid of them or keep them for reference (in my app you can “suspend” cards so that you can still search for them etc, essentially turning them into notes).
If you end up trying Active Recall out please feel free to reach out with any questions or if you just want to chat more about learning!
- Simplified navigation when browsing your content.
- A rich card editor that behaves more like one you’d find in a modern note taking application.
- Search as a first class citizen.
- Dedicated functionality for creating language cards.
- Proper code blocks for people learning software development.
- A way to easily study all my pending items across all folders.
- Improved/seamless synchronization (I’d sometimes get sync failures when using Anki on my phone and on the desktop)
With that said, the application is still very much in its early stages and there are a lot more features that will be arriving over the coming months.
Separately, it's pretty surprising to me to see the very first review by the "VP of Engineering at Shopify" without mentioning the fact that they are an investor in Active Recall. [0] Feels like this is counter to the FTC's rules on endorsements. [1]
[0] https://www.linkedin.com/in/dalmaer/
[1] in particular, § 255.5 of https://www.ftc.gov/sites/default/files/attachments/press-re...
Hey! This will turn into a subscription service at some point but I’m still working out the exact pricing model. Any data that you put into the system however will always be available to you and exportable regardless of if you decide to pay or not. This is mentioned in an email when you sign up but I realize that the communication could be better around this.
> Separately, it's pretty surprising to me to see the very first review by the "VP of Engineering at Shopify" without mentioning the fact that they are an investor in Active Recall. [0] Feels like this is counter to the FTC's rules on endorsements.
Dion was using the product long before he invested and it’s not something that we’ve been trying to explicitly hide (as you saw on his LinkedIn profile). I have updated the review to state that he’s an investor. Thank you for calling this out.
I have a private "notes" repo on GitHub where I keep notes in the issues (the repo itself is empty). Any time I'm trying out a new piece of software I open an issue there, then add notes on the issue comments as I figure things out.
I use GitHub issues because they have excellent backups and they show up on GitHub search - plus there's a really good API which I use to periodically export and backup my notes elsewhere.
If something fits the TIL format, I'll turn my notes into a TIL and publish them on https://til.simonwillison.net - that site uses the markdown format as GitHub issues, so publishing a TIL that started out as an issue comment only takes me a few minutes.
So forgetting is ok, as long as you remember that there was something extra the next time you meet the same problem.
You need to only retain information that can help you derive useful conclusions, and your mind is most likely smarter than you at that.
First, I write code to understand the essence of a piece of software (I have written basic TCP/IP/ARP stacks, a Java compiler, a virtual machine, a database, a neural net library etc). It is a joy to write code for an audience of one when you don't have to worry about error handling!
The second is to add an entry to a diary recapping (in your own words)your own understanding of the topic. In my PhD days, I put everything in a single Word doc; I chose Word because I could draw diagrams in it, search with regexes, and the file was available offline. By the end, the file was thousands of pages long. Here's what I had in it.
1. For every paper or blog post I read, I put the paper's name, the URL, author's name (in BibTex form, but that's not germane to your issue). I captured my understanding of the paper (about half a page) and put together a list of questions for which I needed answers. There were some papers that went right above my head, so that entry was all questions. I had back and forward references to other entries in the form of embedded links. This exercise (it turned out) helped me smoke out scenarios that the paper had not mentioned. That becomes the basis for critiquing and comparing.
2. When my understanding of a topic improved, I would go back and fix some of the outstanding questions, along with a meta comment about what it was about my previous understanding that I couldn't answer the question in the first place.
3. I had design ideas, little bits of code, benchmark results etc all in a single linear order. When my advisor asked what I was up to, I could quickly look up the log of entries.
It is ok to forget; no one can remember it all. Just capturing the essentials and being willing to work from first principles is good enough for most interviewers.
X: I want to be able to recall relevant information when it's needed.
Y: How do I retain long-term memory of technical topics without practising them every day?
To solve X, it is sufficient to be able to reliably rediscover information as it is needed. Putting it into long-term memory stored inside your brain is just one layer of that.
Think of it like a computing architecture, where there are layers of data retention with trade-offs for speed vs reliability vs size. You don't just have a CPU and RAM and a hard drive. A modern computer has various CPU caches, RAM, ssds and rotational drives, possibly combined with offsite storage, cloud storage, tape drives, cdroms and usb sticks.
They all have tradeoffs: RAM is pretty fast, but not as fast as CPU cache, but it holds a lot more than cache, but it doesn't hold as much as an ssd, which in turn is a lot slower than RAM, but not as slow as a hdd, and meanwhile can hold a lot more than RAM, but (generally) not as much as that same hdd. You get the most from your computer by understanding these tradeoffs, not by trying to make your ssd do triple-duty as RAM+ssd+hdd.
Likewise, optimizing your brain doesn't mean trying to force it to keep arbitrary technical data/concepts only in long-term memory. Instead, you are already doing the right thing by keeping the information distributed across the internet, in books, in the brains of people who you teach and learn from, and knowing how to re-derive logically-deducible concepts from axioms, or rediscover data from first principals using experimentation and trial-and-error.
Let your brain do what it does best, which is act as a survival machine optimized for getting mates, finding food, and escaping lions.
Don't hamstring that machine by giving it unrealistic goals like "damnit you simply MUST be able to recall all relevant formulae for orbital mechanics and know how to write down the Navier-Stokes and Black-Scholes equations off the cuff and what is the full instruction set for the ARMv8-A architecture."
Keep refreshing HN to keep up to date, but realize any engineer worth his salt can figure out new tech in a week or two. No need to be an expert.
Sometimes, the fate and knowledge of an engineer can be decided by PM ticket assignment. Side projects are good to keep you sharp, but don’t stress. Once you move to management or team lead, you’ll realize that a competent engineer will always deliver results.
Take notes. Use the Search feature when you need to dig information out.
Memory is not binary: you will have easier or harder access to things in your long-term memory. Those that you access more often, get easier. Those that you build up more complex structures around (understanding concepts underlying those concepts) make it easier to retrieve.
The more you learn and connect information, and generate it (by writing summaries, applying that information, etc.), the stronger that network of information becomes, which means that when you "forget" it (find it hard to retrieve), a quick skim about that topic will bring the whole network of information back to your conscious awareness.
If you want to build better knowledge networks in your brain, you have to (re-)generate that knowledge in different forms. Pick a "node" of that knowledge that's a little fuzzy and learn more about it to clarify it. Draw diagrams of how that knowledge fits together.
Ultimately, constant retrieval of information, with restructuring and clarifying, will make it easier to get back when you need it.
For a lot of programmers, SQL is such a skill. Most programmers suck at SQL because they don't use SQL often enough so they forget the syntax and have to recall it from scratch each time they need it. It's a hidden disability in the software development world.
The answer is that you have to have learned the skill up to a certain level of proficiency before letting it go. You cannot get there by dabbling each time. There's a minimum threshold after which the skill doesn't simply evaporate from memory and recall is fast. But it takes a certain investment of time and effort.
[1] https://www.johndcook.com/blog/2012/08/01/bicycle-skills/
https://activerecall.com is their site
Usage: mash the leftmost pedal to toggle recording the screen; mash the middle pedal to save off the past 30 seconds of my activity (usually along with me verbally narrating what I'm doing, an oddball habit that has paid dividends thanks to this config) and finally, push-to-talk (drops an .mp3 in the appropriate folder; I usually just have it work the same as the leftmost pedal (i.e. saving an .mkv which just so happens to have user audio) because I'm too lazy to bash-script ffmpeg and OBS is already open.
Hardware: I have a tri-pedal set, made by the same manufacturer documented in the 'Vim Clutch' article here on HN (I'm too lazy to go find the URL but it will come up easily in a search.) The pedals are mapped to shortcuts in OBS for the functions described above.
Software: OBS Studio, Gnome desktop (NixOS), and Input Remapper. (More on this last one further down.)
Caveat: This currently only works under Xorg (not Wayland) because Wayland doesn't let OBS grab shortcuts outside its own window (Wayland is such a Karen.) There's a workaround that I'm working on, but it involves an OBS plugin called obs-websocket and a straightforward bit of Python scripting that I have nonetheless yet to finish, as well as Input Remapper (https://github.com/sezanzeb/input-remapper) so Xorg it is for now.
For bonus points, I made the destination folder the 'attachments' folder in my Obsidian vault. Then I can index them by subject, a task that is surprisingly entertaining, as links from topic pages in Obsidian.
Honestly. I don't remember how I thought before. This is the perfect thinking machine.
I do that in Obsidian, a personal knowledge management tool.
I copy/paste contente from the web in interconnected notes, and once the contents are no longer in their original form, but transferred into my own read/write notes, I refactor and reorganise them.
After a few refactors, some contents from different sources are now next to another in the same text. Or disappeared completely because another source has a « better » explanation. You can draw your own things, or add images from the web.
I think this process is a little bit like writing your own Wikipedia article.
The funny part is that when you come back to your notes later, you are astonished by the density of information connected together (just like Wikipedia).
It’s easy to add content to and supports code highlighting, images, files and all the nice things you’d see on wikipedia. And has a beautiful skin (template) with gorgeous typography that’s a joy to read.
I also use it for tech notes, among the other stuff.
I use this for technical webpages that I read, business concepts (e.g. I am reading about `Yule’s Law of Complementarity` or new technical concepts (e.g. NestJS).
This works for me and I stumbled upon this workflow for myself after trying out a lot of other things that did not work for me.
It is highly unlikely I’ll end up in a situation without cell signal, I carry an iPhone, Kindle Paperwhite, and a cellular iPad Mini (256gb) literally everywhere. (Handbags are awesome.)
On the off chance, I’m away from the internet and my kindle doesn’t have a useful book, I have all of Wikipedia and Wikitonary downloaded in Kiwix on my iPad Mini, downloaded PDFs in Dropbox, and a bunch of books in the D&D Beyond app, and videos in Youtube, because why not.
Any thing I need to remember is in Apple Notes or Scrivener. (All my research for stories is in a project file.)
I generally remember the difficult/weird parts that made initial learning slow, everything else I just re-learn as needed from the same sources that I used initially.
One thing to realize is that rapid technology deployment and improvements mean we must adapt our approach to learning and problem solving if we want to be effective.
With the internet and exponential increase in software development tools, the amount of information is massive. And of course, we now have Google and Stackoverflow etc. which are extremely effective ways to solve problems or very quickly refresh your memory on the exact details that are relevant.
If you operate as if those things haven't changed, then you will not be effective.
[1] The Learning Pyramid: https://stephenslighthouse.com/2010/02/26/the-learning-pyram...
It is not a dichotomy and do learning without remembering does not exist unless we have such different definitions of learning that your definition does not imply any structural change on the brain.
This doesn't apply to everyone—if your goal is just the pursuit of knowledge then you can disregard this advice. For me though, the satisfaction and the cementing of knowledge comes from its application. Especially as I get older and realize the ephemeral and fleeting nature of body and mind, it just reinforces the need to be in the moment and cultivate focus on doing rather than optimizing against (or fighting!) my human limitations.
When you learn other things adjacent to the topic at hand, but related in some way, you’re less likely to forget things as you build more and more of a frame of reference around it.
But either way, the key to remembering things for a very long time is spaced repetition. First you need to test your recall after a short period, then after a longer period, etc.
Memorization skills isn’t something you’re inherently good or bad at, or are born with: it’s a skill you can absolutely learn!
if a text file isn't good enough, or there needs to be an image embedded, I use other file types
And a Documents folder containing 'subject' article files. Documents/Programming/Languages/Rust ...
I personally use markdown files in the event I want to publish one of my notes. It also let's me inline images and see them if my editor supports previewing.
Moral of the story, it is possible to get to the point where stuff is in your memory for a couple months comfortably. But beyond that you will have to do a bit of memory maintenance.
1.) Teach things that are really important to me.
2.) Google the rest.
Sadly, I don’t time to learn everything that I want and if I had the time I wouldn’t be able to retain it anyways.
I used to think that Internet would make this easier but it's not the same as running into friends(with diff expertise) on campus everyday. You share your day to day probs. They share theirs. And connections across knowledge maps stored in different chimp brains happen.
in the past i've tried digital apps like anki and quizlet but i'd inevitably get distracted. the point is to come up with a system of review that matches your personality. if you want to remember stuff long term the big supported by science is spaced repetition.
the other system that works but is much harder is to teach someone. pretending to teach a class also seems to work...point is to explain it at level that a new person would understand forces your mind to really delve into the topic.
This may include downloading papers or buying textbooks or equipment to hack (when appropriate).
Strange, but funny feeling.
I’ve been planning to migrate to bear.app for a long time, but have never made the leap.
For example, knowing how to configure your network is useless if you only need to do it once every five years because (sadly) the UI will have changed anyway.
Whenever I learned how to do something new I wrote an instruction for future me with all the key information on how to perform this.
There's no other option for remembering things other than using them.
I've never been sure which one is which - byte or bit
until I had to use them directly