YT YouTube2Text

YouTube2Text → Transcripts → Obsidian Note-Taking System for Academics & Students - Full Tutorial & Demo (Write Papers Faster)

Obsidian Note-Taking System for Academics & Students - Full Tutorial & Demo (Write Papers Faster) — Transcript

by The Effortless Academic · 9,353 words · 1,333 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00So 5 years ago, I decided to start a PhD
0:02in biology. I was 36 at this time and my
0:05knowledge of biology stopped literally
0:07somewhere in high school. But that sent
0:09me on a journey to research everything I
0:12could find about how to learn faster.
0:15I've read the Smart Notes book, the
0:17Second Brain book, researched the Zetle
0:19Custom Method, tried all the digital
0:21tools I could find, and started building
0:24this. That's a digital repository of all
0:27my knowledge over the last four years.
0:30Every paper I've read, every claim I've
0:31highlighted, every question and idea
0:34instantly searchable. So, a little bit
0:36like Wikipedia, but for my own research.
0:39The big difference to systems like the
0:40Zete method or the second brain is that
0:43because every claim is verified with
0:45references, it can be reused
0:48specifically for academic writing, not
0:51just to learn a complex topic. And when
0:53AI came along, I realized that you can
0:55also automate parts of this process with
0:57really good results. A few months ago, I
1:00ended up finishing my PhD 5 months ahead
1:02of schedule and with four first author
1:05papers written. In this video, I'll
1:07share with you exactly how I built this
1:09system, the software you will need, and
1:11the strategy you can take to scale it
1:13over years. And once we have done this,
1:15I will show you how to combine it with
1:17AI. Now, you might ask yourself, why do
1:19I actually need some sort of complex
1:21information management system? And I'll
1:24ask you this, have you ever played
1:26chess? Who wins at chess? It's always
1:28the person with a strategy. So, if you
1:30have ever played against someone who has
1:32even the tiniest bit of strategic
1:34knowledge, they will completely own you
1:36and you have zero chances of winning.
1:39And the PhD is exactly like this. You
1:41need a strategy to win. But our
1:43collective solutions are typically very
1:45different. Check out this nature study
1:46over here. It says that about 78% of
1:50students work more than 40 hours. So in
1:52other words, people grind their way
1:54through a PhD. That's like playing chess
1:56against a grandmaster and losing over
1:58and over and over again. So in the words
2:00of Sansu's Art of War, tactics without
2:03strategy is the noise before defeat. And
2:06since a PhD is about knowledge, you need
2:08a strategy of organizing knowledge. So
2:11let's build one.
2:14Let's keep it as simple as possible and
2:16start with something that everybody is
2:18familiar with, the largest repository of
2:20human knowledge, which is of course
2:22Wikipedia.
2:23Notice something that Wikipedia has
2:25nowhere over here any folders. Nothing
2:28tells you where a page belongs to.
2:31Instead, it's organized around concepts
2:33that are linked to one another. And if I
2:35hover over a single term, then something
2:38will pop up and give me the information
2:39almost immediately, making it very easy
2:42to access. And additionally, every claim
2:45like over here traces back to some sort
2:48of paper or some other source that
2:50proves why this is true. And then at the
2:53bottom, you will find all of these
2:55different books, journals, websites, and
2:57so on cited that make up the foundation
3:00of this article. But also notice what
3:03you will never find on Wikipedia, and
3:05that is a summary of any of these papers
3:08here at the bottom. But now ask yourself
3:10when you read a paper and try to take
3:12notes on it, how do you go about it? Now
3:15chances are you probably start by
3:17summarizing the paper by highlighting
3:20and by doing something with the paper.
3:23Wikipedia has none of that. That's
3:25interesting, right? But that's also
3:27intentional because that's the only way
3:29how we can organize a large body of
3:32knowledge. Instead, Wikipedia is
3:34organized around two main methods. The
3:37first one is links. They connect
3:40different concepts together. What is an
3:42evergreen tree? What are the mortaces
3:44family? What means endemic and so on.
3:47And the second part are of course the
3:48references that separate the speculation
3:51from the fact. Let's take the ideas
3:54behind Wikipedia's organization but
3:57replace the contents with our own
3:59research. So technically each concept on
4:01Wikipedia is a website, right? and you
4:04need a browser like Chrome or Arc to go
4:07through these websites and navigate the
4:09different links. So for our personal
4:11Wikipedia, we're going to use software
4:14called Obsidian as our browser and the
4:16web pages are going to be text files on
4:19the hard drive. This keeps everything on
4:21your computer and allows you to add
4:23additional files like the PDFs of the
4:25papers you're reading, connect your
4:27reference manager, and so so much more.
4:30So here is the plan how we're going to
4:32tackle it. First, I want to show you
4:34briefly how Obsidian works. If you know
4:36that already, feel free to skip ahead.
4:38There's a lot of good Obsidian content
4:40out there already. Second, I will show
4:42you how to organize your academic
4:44reading in Obsidian so that over time
4:47you create something that resembles this
4:49knowledge network of Wikipedia,
4:51including all the PDFs and consisting of
4:53the reusable blocks that you can then
4:55conveniently stitch together to a new
4:57piece of writing. And then in the last
4:59part, I will combine the technology and
5:01the system and show you how it's all
5:02organized. Cuz the biggest challenge in
5:05this entire process is not the
5:07technology obsidian and also not the
5:10system, how you organize notes. That's
5:12pretty simple. It's this mindset shift
5:14from taking sequential notes on papers,
5:16highlighting and all that stuff to
5:18taking topical notes because without
5:21topical notes, we cannot scale our
5:23knowledge repository. [music]
5:27Let's open obsidian over here. It's now
5:29completely empty because I have
5:30absolutely no nodes. If you click this
5:32button over here, new note will be
5:34created and we can call this node
5:35biodiversity and just copy something
5:38from Wikipedia so that we have some meat
5:40on this note. Create another note and
5:43I'll call it tropical forests and also
5:46paste something from Wikipedia. Now we
5:48have two nodes living in our obsidian
5:50vault. So this entire vault of knowledge
5:53just lives on the hard drive and it has
5:55these two notes that I can just open and
5:58preview in another application. If
6:00Obsidian started charging you money
6:02tomorrow to access your information, you
6:04would still have access to it. That's by
6:05the way exactly why I stopped using
6:07Notion and Evernote and all that other
6:08stuff. The first thing you notice when
6:10you look at Obsidian is that there's
6:11absolutely no way to format anything. Of
6:14course, if you double click something
6:15and then rightclick and then say format,
6:18you can bold and itily can strike
6:20through things which makes it a little
6:21bit cumbersome. But the main learning
6:24here is that whenever I bold something
6:27or I italicize it, all I get is a small
6:31change in the actual text. So in that
6:33case, two stars have been added. And
6:36that is what the markdown format does.
6:39It just adds specific characters to
6:41format. But honestly, don't bother
6:44learning that format. It's much easier
6:46to install a single plugin that will
6:48take care of all of that. Just click on
6:51the settings button here. Then go to
6:53community plugins and browse community
6:56plugins and then find one that is called
6:59editing toolbar. It's very very popular.
7:02So a lot of people use it. After
7:04enabling it, you will get this bar here
7:07at the top that allows you to do any
7:09formatting you want. For example, you
7:11could say this is something I want to
7:13highlight. So you highlight it. Or for
7:15example, this one I want to strike
7:17through or make Italy or make bull. And
7:20all of that is just happening for you in
7:22the background. You will still see this
7:24markdown formatting happening and
7:27changing. But you don't need to
7:29necessarily memorize it all the time or
7:31type these characters. And of course,
7:33the key thing that you would want to
7:35have with your notes is headers. So for
7:37instance, this is a definition. So if I
7:39just write definition and then say it's
7:41a header two, it will add these two
7:43hashtags and create header out of that.
7:46The same way I could add another header
7:48risks of biodiversity decline and then
7:53highlight this whole thing, make it into
7:55an H2 header and then add for example a
7:57note of that. So biodiversity is
7:59declining because of climate change.
8:01What I have here is a list. So lists are
8:03very easily done by just typing in a
8:05minus and a space which will create this
8:08bullet point or again you can just use
8:11the editing toolbar and for example
8:13change that to a numbered list. So again
8:15you don't really need to learn all of
8:16that. There is one feature that I think
8:18is super super useful especially for
8:21academic use and that is called call
8:23outs. In the editing toolbar, you would
8:25just click this icon over here and then
8:26the quotation mark. And then for
8:28instance, we could move the definition
8:31into a call out. So I'll just type in
8:36three dots so that we can copy content
8:38later. So see how it created this nice
8:40box here and we can just copy that
8:43entire thing and paste it instead of the
8:45three dots. So now the definition is
8:48tucked away in this box which looks
8:50pretty nice. And a little hack that you
8:52can do with callouts is you can add a
8:55minus here after the square bracket
8:58which will make it collapse by default
9:00and then you can just click to uncolapse
9:03it. So if I create these callouts I can
9:05choose the call out type and there's
9:07quite a big number of callouts. So for
9:09instance if it's a warning I would say
9:12alert something is happening then this
9:15call out will appear in orange. I know
9:19that some people really love callouts
9:21and they even nest them into one
9:22another. So, if you are a person who
9:24really thrives on having different
9:26callouts, then install the call out
9:28manager plugin because this will help
9:31you use callouts much more efficiently
9:33than through the editing toolbar. I
9:36could just briefly show you how I use
9:38callouts. For instance, if you go to any
9:40paper that I have, then typically I have
9:43a summary call out and something called
9:45author's intentions. So that's for me a
9:48way to tell myself why is this paper
9:51written and then what is important to
9:53me. That's less of a summary and more of
9:56something to jog my own memory when I
9:58read that paper. Another use case would
10:00be for example to outsource AI generated
10:03summaries into a call out so that you
10:05can clearly see what is AI generated and
10:08what isn't. The next thing that you need
10:10to know is how to create links in
10:13Obsidian. I'll remove the call out. For
10:15instance, tropical forest is something
10:17that we already have as a note. So we
10:20can instead of just writing the word
10:22create a link to this note over here
10:24which is the note on tropical forest and
10:26we can link these two concepts together.
10:28So to do this all you need to do is type
10:30in the two square brackets like this and
10:34then it will bring up a pop-up dialogue
10:35which allows you to choose which of the
10:37nodes you want to use. So even if I
10:39start writing something like forest,
10:42Obsidian will be smart enough to know
10:43that name is not starting with forest,
10:45but it's just something that's contained
10:47in the name. That makes it a lot easier
10:49to find the nodes that you want to link
10:51to. And if I click that, it creates a
10:53node to tropical forest. And of course,
10:56you can also create links to outside
10:58content. For instance, if I have this
11:01link over here, which is the Wikipedia
11:02article where this definition is from, I
11:05could copy that and say, um, source
11:07Wikipedia,
11:10and then I could highlight the word
11:12Wikipedia and just paste the link that
11:14I've just copied onto this word. What's
11:17Obsidian will do is create a markdown
11:19representation of a link, which is a
11:21single pair of square brackets and some
11:22round brackets. Now, if I click on that,
11:25it will take me to the entry for
11:27biodiversity. This is already 80% of
11:30what you need to know in order to use
11:31Obsidian efficiently. But of course,
11:33sometimes you need more than text in
11:35your notes. You need images just like on
11:37Wikipedia. So if I just go over here for
11:39instance and take that image and copy
11:41that image and I just paste it into my
11:44note, Obsidian will edit to the note. So
11:48the very important part here is that
11:50this image has been now copied into the
11:52vault. That means that everything that
11:54you drop into the vault is copied into
11:56it, which will be important once we
11:58start embedding our PDFs in this image.
12:01If you click on the button here at the
12:03top is essentially also just a link. So
12:06see there are two square brackets and
12:07then the name of the PNG file. The only
12:10difference is that there is an
12:11exclamation mark in front of it. So if
12:13you remove that exclamation mark, all
12:15you have is a link. But if you add this
12:17exclamation mark, then this thing is
12:19embedded. And that by the way also works
12:22for notes. So if I have a note on
12:24tropical forest and I put this
12:26exclamation mark in front of it, it will
12:28embed the entire content of the note
12:30into this note. So just remember this
12:32exclamation mark is something that we
12:34will need. And by the way, you can also
12:37always make this smaller or bigger,
12:39however you need to. And that too will
12:42be reflected in markdown by this pipe
12:46character and then the number of pixels
12:48that goes behind the PNG file. But now
12:50let me show you a slightly more academic
12:52workflow and let's try to embed a PDF
12:55and create a note on a specific source.
12:59So for that I'm going to first create
13:01two folders. One will be our nodes
13:04folder where I'm going to put all of our
13:06notes that we have created and I'm going
13:09to create a sources folder which will be
13:11the folder where all the source notes so
13:13all the PDFs and so on will be stored.
13:16And additionally, I tend to do a support
13:19folder which contains all the images,
13:22embeds, and so on so they're out of the
13:23way. Now, let's create a note for a
13:26paper. I'm going to call this paper
13:302020.
13:32And then just drag and drop the PDF for
13:35this paper. We have a link to the PDF.
13:38The PDF sits here in our vault. We can
13:41put it into the support folder and the
13:42Vanvvic 2020 into the sources folder. So
13:45it's nice and organized. I typically
13:47don't tend to embed PDFs in notes
13:49because that always messes with the
13:51scrolling. If you're scrolling through
13:52the note, then the scroll is going to
13:54get captured when you start scrolling
13:55through the PDF. So it's not really
13:57useful. Instead, I will remove that
13:59embed. And when I read a paper, then I
14:02would for instance just open this PDF
14:05separately. And you can just pull it
14:08over here to the side, which creates a
14:10split view. So you can have your source
14:12note over here where you can take notes
14:14and here on the right side you will
14:16start reading the PDF and making
14:18annotations and so on. Now let's go to
14:20the very top of the file and type in
14:22three minuses 1 2 and three. This will
14:25open the properties dialogue and you can
14:28start adding different properties. For
14:30instance, author would be my first
14:33property vict
14:36enter the journal is PNAS.
14:41So, but instead of typing PNAS, which
14:44would be just text, you can add links to
14:46the properties. So, if I just type in
14:49two square brackets PNAS, this creates a
14:52link to a node that doesn't exist yet.
14:54But the moment you click it, this note
14:56will appear and will be created here in
14:59your folder. The main idea of why we
15:02would have notes on journals is that
15:04later we can just pull all the papers
15:06that are published in this journal and
15:07put them into this note which sometimes
15:09can be useful for understanding where
15:12you want to publish your own work. Go
15:14back to our Vanvvic 2020. And you can do
15:16that by the way with authors as well. So
15:19I could use two square brackets, type in
15:22the name of this author, type in enter,
15:24and that would add the first author also
15:27as a link. I would maybe add one more
15:29property, which is the title. And then
15:31we're just going to copy this title over
15:33here over time. Of course, you will have
15:35many more properties that you can use
15:36for your own research like tags or
15:38categories or whatever else you're doing
15:40that will eventually even replace the
15:42tero. But let me show you why I'm adding
15:45these properties by adding another paper
15:47and then see how we can pull them
15:48together. So I'll create a new node
15:52urban 2024
15:55and we're going to pull similarly a
15:57paper over here another paper on climate
16:00change. Type in the three dashes author
16:05arc urban
16:07journal science
16:10title.
16:12So we have now created basically two
16:14nodes on the different sources. They
16:17have some properties and we might want
16:19to try to bundle all the papers in our
16:22vault. That brings us to one of the most
16:26important and underrated features in
16:27obsidian which is bases. A lot of people
16:30consider basis quite complicated and too
16:33technical to be part of a standard
16:35workflow. I think it's the opposite.
16:37Bases are the absolute foundation of
16:40using Obsidian in an academic or any
16:43sort of structured way. Bases are not
16:46that different from an Excel file. So in
16:48an Excel file you have rows and you have
16:51columns. Then you can add filters for
16:53example to filter your rows by the value
16:56of the column. That's exactly what a
16:58base does. But instead of rows we will
17:00have nodes and instead of columns we
17:02have the properties that these nodes
17:04have. So let's create a new base by
17:06using the command plet. Obsidian has
17:10this wonderful feature. If you click
17:11command P or control P or just press
17:14this button over here, you come to a
17:17dialogue that allows you to do anything
17:19Obsidian can do basically. So if I type
17:22in bases, then it gives me the ability
17:25to create a new base. Let's create a new
17:27base and call it papers. By default,
17:31this base just displays all the files in
17:34our wallet. That is just how bases work.
17:36They take all of the files and then
17:38you're going to have to create filters
17:40and structure to display the files that
17:43you want. For instance, in our case, we
17:45just want to display the files in the
17:48sources folder. To see in which folder
17:50for something is, we can, for example,
17:52go to properties and then click folder.
17:54What we are interested in is basically
17:56everything that's in the sources folder.
17:58So if I sort on that then you can really
18:00see where things are. To filter you just
18:02click filter and then there's a very
18:05convenient drag and drop dialogue. So if
18:06I say folder is and Obsidian is already
18:11smart enough to suggest the folders that
18:12exist for instance sources then this is
18:15what is going to be displayed. The
18:17folder property itself at this point we
18:19don't really need. So we can take it out
18:22but we can add the author the journal
18:24the title and so on. So I can even
18:26rearrange maybe a little bit. The cool
18:29thing about it is that we can now use
18:32this preview feature that Obsidian has.
18:35If you hit command and you hover over
18:38something, it will give you the preview
18:40of this note. And if you again hover
18:44over links in this note, it will give
18:46you the PDF. for example in our case
18:48that's why I always add the PDF link
18:50somewhere to the top so that I can
18:52always drill down to the specific paper
18:57so that's of course not very impressive
18:58if you only have two papers let me
19:00instead show you how that looks once you
19:02progress a little bit here is a base on
19:05the papers that are related to a
19:08specific project that I'm working on
19:11which is essentially a paper and it
19:13creates a base which lists the papers by
19:17their rating and there is a summary that
19:19I have written myself about that. So
19:21these are the foundational works and in
19:23the same way you can also create papers
19:26that you actually want to read. So here
19:28is a list of papers that are a reading
19:31list. So that of course features the
19:33title and maybe the intent why I want to
19:35read this. Here's an example of a larger
19:38vase that contains basically all the
19:40papers in my vault sorted or ranked by
19:44different years or properties. For
19:46instance, I can have all the papers that
19:49released in 2025 and then scroll down
19:52and then find all the papers released in
19:542024. You can create quite elaborate
19:57formulas so that you have all these
19:58different icons appearing. So bases are
20:01incredibly flexible when it comes to
20:04organizing your knowledge. And as you
20:06maybe can already feel like a base like
20:08this can easily replace Zotterero for
20:11you because Zotterero is just terrible
20:13clunky UI. I don't like Zotterero. And
20:16here I have full flexibility. What is in
20:18the rows? What is in the columns? How it
20:20is displayed? I can immediately jump
20:21into the single paper, immediately jump
20:23into the PDF. That's so much faster and
20:26more dignified than using Zero in my
20:28opinion. Don't shy away from bases. And
20:31I have links to two blog posts that I
20:34made about bases and how you can learn
20:36to use them. So armed with this first
20:39setup, let's now move on to the second
20:42part where we're going to try to
20:44understand how we are going to structure
20:46these notes once we have hundreds or
20:48thousands of notes because that is where
20:50the real challenge lies is to keep a
20:53structure that scales with you and where
20:55you can find your notes. That's I think
20:57where I made the most mistakes with and
20:59I hope that the next part will be really
21:01useful for that.
21:05Here's where most people go wrong. To
21:08think note takingaking means summarizing
21:10a paper or any sort of source, but
21:13summaries aren't that useful. For
21:15papers, you typically have an abstract
21:17where someone who knows the paper far
21:19better than you spend hours and hours
21:22summarizing it. So summarizing is just
21:24compressing knowledge. So it's easily
21:27transportable from one person to
21:28another. Think of it as a suitcase.
21:31Compresses your clothes beautifully. But
21:33try to find any specific item and it
21:36gets really really difficult because you
21:38have to open the entire thing. You have
21:40to look. You have to rumage through it.
21:41By this time you kind of already forgot
21:43what you're looking for and so on. So
21:45what you actually want is a wardrobe.
21:48And that requires a mindset shift in how
21:51you organize your clothing or
21:54information in that case. So instead of
21:56taking notes on sources and PDFs, we
21:59first understand which topics these
22:01sources deal with and then we extract
22:04the claims a source makes on these
22:07topics into the respective nodes. After
22:10experimenting quite a while, I think
22:12there are only three types of nodes that
22:14you need to ensure that the system
22:16works. First, you need source notes
22:18which contain everything someone else is
22:21saying. They're not really mutable. So
22:23you just write them and you forget about
22:25them. And then the topic notes which
22:28contain all the claims you extract from
22:31these papers. They are subject to your
22:33interpretation but not speculation.
22:37Every claim here in the topic note links
22:39back to the source note. Number three is
22:43the argument or thinking notes which
22:46contain your own thoughts, your own
22:48ideas, your own question. So this is a
22:51clear separation between what they said,
22:53what you understood they said and what
22:56you said. So sources, topics and
22:59arguments is what enables rigorous
23:02research in my opinion. And we have
23:04already started in our obsidian vaults
23:06to have a source folder and a notes
23:09folder. And we have already added this
23:12first paper into our source folder.
23:14Let's try to look at this paper and just
23:17try to read what it's happening. so that
23:20we understand how these source notes
23:22link back to the topic notes. So for
23:25instance, this passage over here says
23:27that evidence of climate change impacts
23:29on biodiversity is accumulating. But a
23:32growing literature also reveals highly
23:35valuable rates and even directions of
23:37these responses. That is something that
23:40relates to the concept of biodiversity.
23:42And we have already a note on
23:45biodiversity. So let's go to this note
23:48and basically try to extract that claim
23:50from this paper over here into
23:53biodiversity.
23:55For instance, we could say rates of
23:57biodiversity change are highly variable.
24:00So it is a compression of that claim.
24:03And to make sure that it's not a claim
24:05that I'm making but the paper is making,
24:07I'm going to add a link to Venvic 2020.
24:11So this ensures that whenever I read
24:14something on biodiversity and I think oh
24:17that's a really good point that I want
24:18to reuse somewhere I know where to
24:20double check namely in this Vanvvic 2020
24:23paper. And remember in Obsidian you
24:26could always hover something with your
24:29command key and then it will pop up the
24:31contents of the source note which at
24:33this time only contains some properties
24:35but is essentially empty and it contains
24:38the link to the PDF which helps me open
24:40up the PDF. So right now the PDF is of
24:42course open here on the right but if you
24:44have hundreds of nodes this becomes
24:46really really useful way of actually
24:49seeing the raw source immediately in a
24:51few steps. And there's a small hack I
24:54would say that I'm using. So if I copy
24:56the title of the paper and go to this
24:59PDF link and after the PDF I'm going to
25:02type a pipe character and paste that
25:05title there. What this does is it will
25:09display the title but actually link to
25:12the PDF itself in the background. So
25:15this allows me to hide the PDF behind
25:17the title. And if I just add one hashtag
25:20in front of it, then it gets nicely
25:23formatted as the title. And here I could
25:25say, okay, well, summary of the paper or
25:27whatever else I want to do like author's
25:30intentions is something that I often try
25:33to extract,
25:35core theory and so on. So whatever else
25:38you would want to add into your source
25:40note. Let's do this exercise one more
25:43time. So here for example they say we
25:45propose that this complexity arises
25:47because biological processes do not
25:50respond to climate in direct and linear
25:52ways. The statement here might lead us
25:55to create a new node. So I'll create a
25:58new node and call this node biological
26:01processes.
26:03I will put it for now manually into the
26:05notes folder and just write down that
26:10responses
26:11to climate are not linear and again add
26:16a link to this Vanvvic paper. But in
26:19addition to that we can go to our
26:22existing note and say reason
26:26is that biological processes
26:30have nonlinear responses.
26:33So what we did now in addition is that
26:36we linked this biodiversity concept to
26:38the concept of biological processes
26:40where we will have all the other
26:42information on that topic of biological
26:45processes and that's already pretty much
26:47resembling what you have seen in
26:49Wikipedia plus with a little bit more
26:52rigor when it comes to sourcing these
26:54papers and because we have added this
26:56title and the PDF link it gets much
26:59nicer now to look into the PDF. Obsidian
27:03has this wonderful feature where you can
27:05link directly into the PDF. So if I just
27:08highlight something, for instance, this
27:10sentence here says in herbaceous plant
27:13communities, warmer or wetter conditions
27:15lead to increased biomass and height. I
27:19can rightclick it and copy this link to
27:22selection. And then let's say this
27:25increased biomass and height is a
27:27biological process. So it would go into
27:28this note over here. And then I'm going
27:30to paste that link. So this link gives
27:34me a link to the PDF directly. So if I
27:37hover over that instead of going to the
27:39source note, it takes me directly to the
27:42sentence in the PDF. That is a really
27:45great way of double-checking anything
27:47because for that particular highlight,
27:50we say weather or warmer conditions
27:52increase biomass and we link to this
27:54Vanic 2020 paper. But in reality, it's
27:58not that paper that's making this claim.
28:00It's paper number 24 and 25 in the
28:04reference list to this paper. So if I
28:06were to site weather, warmer conditions
28:09increase biomass and then say it's vanic
28:112020 that say this, somebody would look
28:13it up and say, "Hey, that paper has
28:15nothing to do with what you're saying
28:16here. It's just a sentence in the
28:18introduction." So this is a red flag and
28:21that's why if you have a link to the
28:23PDF, you can immediately check whether
28:25or not that's true and then go to source
28:2724 and 25 and see who actually has made
28:31this claim. Here the convention that I
28:33have for myself is the so-called in
28:35link. So I'll replace that anchor text
28:38with in and then I will add the link to
28:43the source note itself. So now I have
28:45two links per claim. One of which goes
28:50to the PDF only to double check whether
28:52the claim is true and one of them goes
28:55to the source note itself which allows
28:58it to connect to all the other notes
29:00because see what's happening. If I open
29:01this Vanwick note and then I go and
29:04click those three dots over here and
29:06open the back links in the document I
29:08will have all the linked mentions. So
29:11all the other notes that link back to
29:14this paper and that tells me to this
29:17paper that it deals with one
29:18biodiversity, two biological processes
29:21and of course as you keep taking notes
29:23over the entire paper you'll have 10
29:25different concepts and you will
29:27immediately ground that particular
29:30source with the different concepts it
29:32deals with. And it goes also the other
29:34way around. If I go to biological
29:36processes and you imagine there's like a
29:38hundred more of such notes, it will give
29:40me a list of papers that deal with this
29:42topic. Over time, the topic notes become
29:45like a mini literature review on that
29:47particular narrow topic, while the
29:50source notes become gateways to all the
29:52different topics that they deal with. Of
29:54course, it's a little bit cumbersome to
29:56always have an in and a ling and have to
30:00set it up manually basically every time.
30:02So, this can be automated by using a
30:04plug-in called PDF++.
30:06It will also add highlights directly to
30:08the PDFs, so you can see where something
30:11is linked and what is important and go
30:13back and forth between PDF and notes.
30:15It's a little bit of a thing to set up
30:18PDF++ and I've added a tutorial blog
30:21post so that you can read it up and just
30:23set it up for yourself once you are
30:26starting with this system. There's maybe
30:28one addition and when we go to our
30:31source node over here, what would you
30:33add to this source note, right? It
30:36essentially is just acting as an anchor.
30:39Well, I use source notes primarily to
30:42contain the primary source material,
30:44which for papers is the PDF itself. But
30:47sometimes you would want to extract
30:50specific figures because it really is
30:52tightly connected with your source note
30:54and not really related to everything
30:56else. It just emphasizes the result or
30:58the ideas that this paper presents. So
31:00it belongs into the source note. Hit
31:03command shift 4 on a Mac and highlight
31:06that. Then it creates an image and you
31:09can just drag and drop that image over
31:11here and you have that figure screenshot
31:14in here and can start writing your notes
31:16and say okay well that's the main idea
31:18of the paper is to present something
31:20that's like this cuz this has the
31:23benefit that if you are looking at your
31:26topic note and you open the source note
31:28you immediately get to see the figure
31:31and some interpretation that might be
31:34important in the context of this topic
31:36note.
31:37But there's of course the case where you
31:39don't have a PDF where you're taking
31:41notes on let's say a website. Here is an
31:45example. It's also a biodiversity
31:48article and that's probably something
31:50that might contribute to whatever I'm
31:53learning or writing. And Obsidian has a
31:55really great way of importing such
31:57sources that are not PDFs using an
32:00extension called web clipper. So you
32:02create templates and each template
32:04imports the source somehow differently.
32:07You select the vault you want to have
32:10the folder where you want to import it
32:11and then just add it to obsidian. This
32:13imported now the entire page with all
32:16the images and with all the text and all
32:18the links that it included. So the
32:20source note now is not linking to the
32:23PDF but is containing all the
32:25information itself. In any way possible,
32:28you should always aim to have your
32:30source note links to either the PDF or
32:32the text or whatever the raw material is
32:35because that's the only way how you can
32:36doublech check whatever claims you have
32:38extracted in your topic notes and trust
32:41me you will make mistakes when you
32:44extract these claims and over time
32:45you're just like what was I thinking cuz
32:48your thinking develops and you get
32:50better as a scientist and so you see
32:53other things in the same words. Before
32:56we continue, let me just show you how
32:58this looks once you have done this for a
33:01couple months or years. Here's one of
33:04the first topic notes that I have taken
33:06in my vault. And it works exactly how
33:09I've shown you. There's one thing that I
33:11think is very important to see, and that
33:13is that all of these notes are pretty
33:15short. So, I call them atomic sentences
33:18because they're sentences that are
33:20basically minimalistic and that you
33:22don't need to add anything else to. And
33:24what you can do now is you can treat
33:26them as little jigsaw puzzle pieces that
33:28you can use to form an entire paragraph
33:31of knowledge. You can do it manually by
33:33just thinking how could I combine all
33:35these 25 bullet points into a paragraph
33:38or you can ask AI to do that with a
33:40little bit of extra prompting. You get a
33:42paragraph that looks pretty much exactly
33:45like you would want an academic
33:47paragraph to start off. The results will
33:50never be perfect, but they will be a
33:52really good first draft for a paragraph
33:54on forest, let's say for your
33:56introduction section, and it will
33:58already contain all of the links to the
34:01right references.
34:03The power of these atomic sentences is
34:06that over time, they write the papers
34:09for you. That's why the system makes
34:11writing papers so much easier and
34:13faster. And double-checking what you're
34:16actually writing also becomes easier
34:18because you have the notes to the source
34:21notes and in the source notes you have
34:22the PDFs and in the PDFs you can always
34:24double check all the claims and see all
34:26your highlights and so on. There is of
34:28course one thing that we have to ask and
34:30that is is this type of writing ethical?
34:34Well, I have a story about that and I
34:37created a post where I described the
34:39system like this and that almost got me
34:41expelled from the university. It was a
34:44huge storm to say the least. And
34:47here's a video of how that whole thing
34:49went. Watch the video and make up your
34:52own opinion. There's one last thing that
34:54we need for our note-taking system. And
34:56these are argument notes. Because so far
35:00we have sources made by someone else. We
35:02have topic notes which contain our
35:04interpretation of the sources. But where
35:06do we store our actual ideas, our
35:08thoughts, opinions, questions, and so
35:10on. So that's what argument nodes are
35:12for. You can also call them thinking
35:14nodes. They are separate from all the
35:17other nodes and they contain pure
35:20speculation so to speak purely you
35:23nobody else. And this separation is
35:25essential for this academic note-taking
35:28system because you need to understand
35:30what claims others have made and what
35:32your speculation is. Here are a few
35:34examples that I have in my own world.
35:36For instance, this note over here is
35:38asking the question, does raising have
35:40an effect on range shits? And what I did
35:42is I added an email thread that I had
35:45with somebody from the scientific
35:48community on that. They had really good
35:49points and led me to good papers and so
35:52on and so on. Cuz here at the bottom
35:54again with the linked mentions, I can
35:56see where it was used. And if I wanted
35:58to explore that question further, I
36:00would already know where to start
36:02because this paper linked to this
36:04question. And then over time that
36:05question note grew by the email exchange
36:08with John Campbell. Another way of
36:10having an argument note is over here. I
36:12have just written up something. I have
36:14added the sources. I have added bullet
36:17point introductions, my thesis aims, the
36:19chapter outlines and all the things. So
36:21it's completely me. That's why it is an
36:24argument note. The note-taking system in
36:26its simplest form consists basically of
36:28three types of nodes. The source nodes
36:30that are immutable and written by
36:32others, the topic notes which are claims
36:34we extract and then interpret and the
36:37arguments nodes which contain our own
36:39thinking, our questions, our struggles,
36:41ideas and so on. Technically, we can
36:44have them live in separate folders and
36:46this allows us to use a variety of
36:48plugins to style them differently, have
36:50different colors and so on. And of
36:52course, there are billion upgrades that
36:56you can add to this system. And you
36:58might have noticed already in my vault
36:59there, there's a lot of stuff going on
37:01that is not standard obsidian. And the
37:04main purpose of that is to make this
37:06system more frictionless. Cuz
37:08effectively what I have outlined so far
37:10in this video is the concept of how to
37:13build a wheel. Yeah. But wheels built a
37:15thousand years ago are very different
37:17from wheels built today with all the
37:19ball bearings, the rubber coating, the
37:21suspension, and so on. So these are
37:24improvements that you will over time
37:26absolutely discover by yourself if you
37:28play around with Obsidian. You can check
37:30out my blog for more tutorials. I've
37:33been writing about it for many years.
37:35You can, of course, subscribe to the
37:36newsletter and get a weekly tutorial. Or
37:39if you want a really deep dive, I have
37:41an entire course recorded on how to
37:43build every single aspect of my vault
37:45and contains about 8 hours of video in
37:47small chunks and adds you to the
37:49community where you can ask questions.
37:52But that's all just a shortcut because
37:54you figure all of these things out on
37:56your own. Don't stress about it. Don't
37:58try to learn everything at once. This
38:00will just overwhelm you. Obsidian is
38:02complicated as hell if you start out
38:05with it for the first time. To avoid
38:06this overwhelm, let me show you now in
38:09part three exactly how I take notes with
38:11the simple system that we have built so
38:13far without any of this extra coloring
38:16and plugins and so on.
38:20For this demo, I'm going to use the
38:21article that we clipped because you
38:23don't need any domain knowledge to
38:25understand it. Let's first clean up
38:27Obsidian. Let's close all the other
38:30tabs. Close everything in here. But I'll
38:32just leave a note open here on
38:35biodiversity. The way that I usually do
38:37it is I have my reading on the left and
38:39I have my notes on the right and we can
38:41for now completely forget about the
38:43sidebar. Let's start reading. A small
38:45boy holds enthusiastically in his
38:46fishing rod. The lion flies up and the
38:48needle spin fish strikes him on the eye.
38:50Desperate to stay outdoors.
38:53So at this point I might already start
38:55thinking, okay, maybe that's all just
38:56fluff. So we can just skim through it
38:59and see. Oh, okay. Well, the father of
39:01biodiversity is being introduced here
39:04and the man is EO Wilson, the eminent
39:06American biologist. I can just reformat
39:10it a little bit so that I can highlight
39:13that single sentence. In the next
39:15sentence, we have in his book, the
39:16diversity of life described biodiversity
39:19as an assemblage that has eaten the
39:21storms, folded them into genes, created
39:23the world that created us. So we could
39:26go ahead and say okay well that guy is
39:28somebody who defined biodiversity. So we
39:32could definitely write that down. Maybe
39:34we would say defined by EO
39:38Wilson and Eo Wilson because it's a
39:41person he might write books. We might
39:44later read that book the diversity of
39:46life. So I will create a link out of
39:49that just by placing the two square
39:52brackets. And maybe I would add in his
39:54book the diversity of life. The
39:57diversity of life itself will be a
39:59source note eventually. So I would also
40:02make that into a link. So let's first
40:04click on the diversity of life and check
40:08that this node is sitting in our sources
40:11folder because the diversity of life
40:12will eventually be a book. We can
40:15introduce a few properties. For
40:17instance, I could say the type of this
40:19is book. I would then add maybe another
40:22property which is authors. And you see
40:24how Obsidian already suggests the two
40:27authors that we have used in this wall.
40:30One is elizabeth bo who is the author of
40:33this article and the other one is this
40:35van paper that we've been working on. So
40:38now we will add a third author which
40:41will be EO Wilson. And as you see,
40:44Obsidian already knows of the EO Wilson
40:47note even though I haven't created it
40:49yet. So I will just link it and because
40:53that is something that I want maybe to
40:56read in the future, I could add to
40:57myself read it to understand the origins
41:01of the biodiversity subject. And of
41:04course biodiversity itself is a topic
41:06that we already have. So I can place it
41:08in square brackets to have the link to
41:10biodiversity over here.
41:12So the next thing is in his book the
41:14diversity of life and we could actually
41:17immediately add a link to this book
41:20because we already have that note. He
41:22describes the biodiversity and so on and
41:25so on. We tend to think of biodiversity
41:28as a landscape teeming jungles and coral
41:30reefs. It's destruction manifesting as
41:33forest clearance and species extinction.
41:36However, these images don't capture the
41:38full significance of the equilibrium
41:40that Wilson described. Biodiversity is
41:43not just the abundance of life on Earth.
41:45Rather, it is what maintains the
41:47resiliency and flexibility of the
41:49environment as a whole. That's an
41:51important passage to me as a researcher.
41:53Maybe if I'm studying the concept of
41:55biodiversity, all the rest is maybe not
41:58so important. So again I'm just going to
42:00highlight that part and try to create a
42:04note somewhere in my topic note
42:06collection here under biodiversity we
42:08could start collecting this idea but
42:11even at this point the note for
42:13biodiversity becomes kind of a of a dump
42:17because there's on the one hand there's
42:18a risks of biodiversity decline on the
42:20other hand there's a definition then
42:23there's kind of some sort of effects of
42:24biodiversity so instead I'm going to
42:27create a new node impact effects of
42:30biioiversity
42:32decline. And we are just going to copy
42:36all of that stuff that we had here
42:37previously. Maybe remove that old stuff.
42:40So I'm going to copy that into this new
42:42node.
42:43Add link to biodiversity so that we
42:46create this knowledge network. And we
42:49can now finally add this claim. So
42:52decline leads to diminished resilience.
42:58and flexibility.
43:00Notice that I'm adding my interpretation
43:02of this because here I'm speaking about
43:04biodiversity decline while this source
43:07just speaks of biodiversity itself.
43:10That's what I mean by claims that are
43:12interpreted by you into the context of
43:15your knowledge. From here, we of course
43:18need a backlink to all of this stuff.
43:21add a link to this source. But instead
43:26of just linking to the entire note,
43:27which again will be very hard to verify,
43:29I want to link to that particular block.
43:32Obsidian has that feature by pressing
43:34this carrot key. And here I can select
43:37any single block that I want from all
43:40the blocks that are in the article. So I
43:43will just add a link to this block. This
43:45adds this cryptic character combination
43:47which also appears after the block. But
43:50that makes it easy for me to command
43:53hover this and only see the block. And
43:55the highlight itself makes it clear what
43:58I'm referring to when I'm linking to
44:00this block. So it's the same principle
44:03as with the PDF, but now twist it
44:05slightly differently. And because I
44:07don't want the link to look so ugly, I
44:09could just type in the pipe character
44:12and say it's Bokes 2018. So Bes 2018 is
44:17the name of the author and Obsidian was
44:20smart enough to suggest this to me
44:23because Bes 2018 is stored as an alias
44:27for this node and that's another cool
44:29feature that you can use to just find
44:31things as aliases because aliases help
44:34you linked here a little bit easier.
44:37Let's continue our reading. The global
44:38convention on biological diversity
44:40defines its subject as it variability
44:42among living organisms at three
44:44different levels within species between
44:46species and of ecosystems. So we already
44:49have a note on this. I'm just going to
44:51open this over here and put it and say
44:55defines
44:56biodiversity
44:58as we can even copy this part over here.
45:02add a quote and then put it in and say
45:08source books 2018 which will link it
45:12over here. In recent history, many
45:14species have been reduced to far smaller
45:17numbers. Pre-Colus, 25 million bison
45:20roamed in the plains of North America,
45:21but by the late 1800s, fewer than 100
45:25remained in the wild. That's definitely
45:27a fact that might be worth saving. So,
45:29let's highlight this and see now we have
45:32all these four links here. It would be
45:35really nice to be able to switch between
45:37them a little bit faster. So, Obsidian
45:40offers the possibility to stack those
45:42tabs. If you just click this icon here
45:44and you stack the tabs, then you can
45:47just scroll through them horizontally.
45:48Yeah, if you're using a trackpad or an
45:51laptop, that's extremely convenient
45:53because this allows you now to have one
45:54source and like all these different
45:56topic notes stacked on the side as you
45:59take this notes because where would this
46:02fact go? It could go to general
46:05biodiversity. It could go to impacts of
46:07biodiversity decline to but it kind of
46:10needs its own node. So maybe examples of
46:13biodiversity declines. So let's make a
46:15new node. examples of biodiversity
46:19decline.
46:21And then we could say American bison 25
46:26million goes to 100 in 1880.
46:31And again, we can go to boats 2018
46:36and link to this block with the bison.
46:39If I just type in bison, it will give me
46:41all the blocks that contain the word
46:42bison and it's only this one. So we'll
46:45just add bison and now we can always
46:47double check that particular fact and so
46:50on. So I think you get the point that we
46:52are slowly building a network of ideas
46:55and that we're always reading in the
46:57same way. We are highlighting something
46:59on the left and we're adding those notes
47:00on the right. And by the end of this
47:02process what you're having is a
47:04knowledge network. So if you open the
47:06graph view here in obsidian
47:10it will show you all the different nodes
47:12that we have. Here's our central node of
47:14biodiversity and then all the connected
47:16other nodes over here. Now that you have
47:19that, you could add a few groups. And
47:21one group could be that the path matches
47:24sources, then it wants to be in red. And
47:28another group could be that the path
47:30matches notes, then we want this to be
47:33in blue.
47:36So that helps us distinguish what is a
47:39source and what is a topic node and
47:41helps us show this knowledge network
47:43over time. That just took us I don't
47:46know a couple minutes. So let me show
47:48you how that looks after 5 years. Here
47:50is my vault again. And if you just open
47:53the graph view, there's thousands of
47:54nodes. Maybe not thousands but hundreds
47:56of nodes for sure. And they're also
47:58coded. The green ones are topics and the
48:01blue ones are sources. So very similar.
48:04And we can jump into each individual
48:06node. We can see all of the different
48:07connections. So the graph view in itself
48:10is just purely visually impressive, but
48:14it's not very useful to navigate around.
48:16It just visualize what you have actually
48:20created as you went and as you created
48:22all these links. Over time, this
48:24knowledge network becomes an extension
48:26of your mind. Because every time you
48:28take notes, you meditate on the thinking
48:31process and you document what you have
48:33been thinking as you have been reading
48:35all the sources and the connections you
48:37have made. And the knowledge network is
48:39a reflection of this. And moreover,
48:41jumping into any of these notes allows
48:43you to immediately discover tons and
48:45tons of different other relevant
48:47connections, be it to other topics or be
48:49it to sources.
48:51So, I hope this demo makes sense to you
48:53and you see how powerful over time that
48:56simple note-taking system is and the
48:58severe mindset shift that we need, which
49:00is to take notes on topics.
49:05We set out to create a personal
49:07Wikipedia for our research that would
49:10allow us to write papers faster, find
49:12everything that we have ever read, and
49:15also to generate new research ideas. The
49:18key to achieving this was to create
49:19so-called topic notes consisting of
49:22atomic sentences which naturally over
49:25time became narrow literature review
49:28clusters which are super easy to find
49:31and which you can update as your
49:33knowledge grows on a topic. And because
49:35we took those notes as claims from other
49:39papers using these small atomic
49:41sentences, we could use them or reuse
49:44them to draft our paper much faster. and
49:47even have AI help us in this process.
49:50And the last part where argument notes
49:52where we just put our own understanding
49:55of the subject matter, our speculation
49:57and so on. This trinity of source, topic
49:59and argument is what creates a system
50:03that tracks every claim to its source,
50:06allows you to express yourself and at
50:07the same time grows with you over years
50:10and years and maybe even decades to
50:12come. In the beginning, I also said that
50:15generating ideas is a side effect of
50:17using that system. So, let me explain
50:20this to you. We start here and explore a
50:23note on fossil fuels for example. It
50:25links to another concept which is maybe
50:28climate change. Just reading this and
50:30navigating this climate change note, we
50:32might be finding that one of the impacts
50:35of climate change on the planet is
50:37melting ice sheets. That's another topic
50:40note. And this note is then again maybe
50:42linked to your favorite subject,
50:44penguins. So a penguin note in itself
50:47might contain even a reference to a
50:49paper that you read just for fun a
50:51couple years back about declining
50:53penguin populations in Antarctica. What
50:55you did here is actually not trivial
50:57because you synthesized a connection
51:00between declining penguin populations
51:02and the newest fossil fuel emission
51:04reports. You can now consider creating a
51:07new idea or question note and ask
51:10yourself how are penguin populations
51:12impacted by the new fossil fuel emission
51:16projections or something like that.
51:19There are a myriad of tools that can
51:21help you find related papers to these
51:23two and get you started on the
51:24literature review. So the idea itself
51:27was generated actually by just walking
51:30this knowledge graph. We didn't do some
51:32sort of idea generation ritual or
51:34brainstorming. It just happened by
51:37navigating the network. And that is why
51:40the idea generation in this note-taking
51:43system is a side product of a well
51:46organized structure. I see it a little
51:49bit like a garden. Your notes are
51:51plants. You saw them. You create the
51:54soil for them. You link them. You curate
51:56them in a certain way. But from there,
51:58you just have to wait. Yeah. and let
52:01your knowledge network grow over time.
52:04If you did everything right, then the
52:06harvesting part, the idea generation,
52:09the paper writing, that's actually the
52:11fun part. For most people, it's actually
52:13exactly the opposite because they are,
52:15so to speak, hunters of information.
52:17Whenever they get hungry and they have
52:19to write a paper, then they go hunt for
52:22new ideas. They go hunt for how to write
52:24all of that. And that takes a lot of
52:26effort because you have to find all the
52:28papers again, try to upload all of that
52:30into your mind and then get to the
52:32writing part. We are farmers of
52:35knowledge. We planted this knowledge
52:37maybe years ago into the soil of the
52:40knowledge network. And when the time
52:42came to write a paper, all we needed to
52:44do is pluck different notes and create a
52:48paper out of it. So the apple was
52:50growing there on the tree. All we had to
52:53do is grab and bite it. But keep in mind
52:55that once you have eaten your first
52:56apple, you are never going back to this
52:59pen and paper notes and PDF highlights.
53:01Digital note-taking is literally a
53:03one-way street. So guys, this video got
53:07astonishingly long. I'm really glad you
53:09stuck around for all of the duration. So
53:12please post your questions and comments
53:14below and I'm really glad to answer them
53:16or guide you to one of the blog posts
53:18that I've written in the last couple
53:20years. And if you want to learn more,
53:22then please sign up to the effortless
53:24academic newsletter. I will send you a
53:26free 21-day course to walk you through
53:29setting up this system with all its many
53:32components and also add a few more
53:34things we didn't have time to talk about
53:36like project management, mind mapping,
53:38the different plugins and so on. So,
53:40thank you so much for watching and I
53:43hope you make your notes work for you.

About this transcript

This page contains the full transcript of Obsidian Note-Taking System for Academics & Students - Full Tutorial & Demo (Write Papers Faster) by The Effortless Academic, generated from the public captions YouTube serves with the video. The transcript has 9,353 words across 1,333 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.