PodSearch

Cortex

179: The Philosophy of Obsidian, with CEO Steph Ango

 

00:00:00   Hello, welcome back to Cortex and the State of the Workflow series. Today I have the pleasure

00:00:04   of being joined by Steph Anger, the CEO of Obsidian. If you're a long-time listener of this

00:00:10   show, you'll know Obsidian. It has been a part of our workflows. We've used it, recommended it,

00:00:16   debated about it. It is a tool that many of you feel strongly about, so it feels fitting to actually

00:00:23   talk to someone from Obsidian. If you're not aware or you need a refresher, Obsidian is one of

00:00:29   the most popular and powerful tools for writing, thinking, and organizing your notes. And it's

00:00:35   built on a philosophy that is somewhat unusual in the software world today. Your notes in Obsidian

00:00:40   are just files on your computer. They belong to you, not to the app. What makes Steph's story

00:00:47   particularly interesting is how he got to Obsidian. And I want you to hear this story. I also wanted to

00:00:53   have Steph on the show to talk about how he works, how a small team builds a product used by millions

00:00:59   millions of people and the ideas behind it all. So let's get into it.

00:01:03   So Steph, I want to ask you as I ask everyone that comes onto the State of the Workflow series,

00:01:09   what is the most important device that you use to get your work done?

00:01:14   I would say my desktop computer now, but it has been a change because my MacBook Air was for probably five

00:01:24   six years, the main device that I did almost everything on. And I loved that because I could

00:01:29   easily go to my office computer, plug it into a monitor and do it there or sit on the couch and

00:01:35   keep working and have everything be super fluid. But now I'm a lot more of a mixed device type of person.

00:01:42   I'm switching between my phone, my MacBook Air and my Mac mini. But I think nowadays I'm doing most of my work on my Mac mini.

00:01:51   So you are the CEO of Obsidian. How would you describe Obsidian to someone who hasn't used it before?

00:01:57   Well, first of all, CEO is a very highfalutin title for a company with like seven, eight people. So, you know, we're a note taking app where the probably the main differentiator is that it works a little bit like Wikipedia.

00:02:11   So when you're writing, you make links. So I'm talking to you right now. It's say in my diary for today, I was on the Cortex podcast with Mike. And each time I mentioned something, whether it's a person, a place, a podcast, a book, a concept, I can make that a link just the same way that when you're browsing Wikipedia, everything is a link.

00:02:33   And what's nice about that is the structure comes from those links. So rather than organizing your things with folders and tags, which you can certainly also do in Obsidian, if you're that kind of person, you can navigate and browse your ideas through links.

00:02:47   And that's kind of the basic concept. Then you kind of get into some nerdier topics about how the data is stored. You know, we use markdown files, which are very durable.

00:02:56   It will live long beyond the life of Obsidian. Or if you have been using another app, it's very easy to import. There's plugins and themes and we can go there if you'd like. But the kind of basic idea is how do you create, whether it's a knowledge base, a world that you want to invent, or just your own personal journal, a way to connect your ideas.

00:03:18   So we will touch on many more things about Obsidian, the app, I'm sure, throughout our conversation. But I want to kind of talk about you and your context and the role that you're in. You're not a founder of Obsidian. I think that is something, especially with a smaller team, I think people would assume that the CEO was a founder or had been there since the beginning.

00:03:37   But you were known initially, I feel like for making one of the most, if not the most popular Obsidian theme, which was called minimal. So going back to this point, what drew you to Obsidian and then also to create kind of your own theme for the app so it looked and worked the way you wanted it to?

00:03:54   So at the time when Obsidian launched, it was March 2020, which was also an important moment for everyone in the world.

00:04:03   And at that time, I was running my previous startup, Lumi, which was essentially a manufacturing marketplace, a B2B platform.

00:04:12   And I don't think I've ever had a job. And even now, I kind of feel like I don't because it's so fun to work on Obsidian every day.

00:04:19   But I had always been running my own startups. And when Obsidian came along, it fit very naturally into what I had been doing, which is note taking with Markdown and trying my best to use this kind of Wikilink type of format that I described.

00:04:37   At the time, I was using another tool called TiddlyWiki, which is still around. But this idea, now it's become so widespread that I almost feel like going back into what the world looked like 10 years ago seems very archaic.

00:04:50   But all of the Wiki software that existed back then was primarily designed around publishing something like Wikipedia.

00:04:57   So there was like MediaWiki, which is a big old piece of PHP software that runs a lot of Wiki software out there.

00:05:04   But there was a couple of apps, including TiddlyWiki, that were doing it like that, more for personal use.

00:05:10   And I didn't really love how it worked, but it kind of had that basic functionality.

00:05:16   And when Obsidian came out, it instantly just made sense because it was doing a lot of what I was doing with TiddlyWiki, but with a much nicer UI.

00:05:23   You know, the search was good. It had some of these intuitive features that now we're used to seeing in almost every app, like a command palette.

00:05:30   You know, I can quickly get into whatever note I want to.

00:05:33   And so, you know, my old system was like TiddlyWiki with a bunch of CSS and hacks on top of it.

00:05:41   Porting over to Obsidian was super, super simple, even in like the first version that came out.

00:05:46   So the one challenge that I had was for me, the interface felt a little bit not familiar to Mac OS, which is where I primarily do work.

00:05:56   So I just immediately started tweaking it to make it feel a little bit more at home on my computer.

00:06:02   And coming from a design background, that's just something that I had fun with and kind of creating your own personal workspaces is nice.

00:06:10   But I was running a company with 45 people.

00:06:12   And so this was just, you know, a part time hobby, like on the side.

00:06:16   But I was using Obsidian itself to take my notes for work, meeting notes, my journaling, keeping track of my life, all of those things.

00:06:25   So it just came out of a necessity to make the app a little bit nicer for me.

00:06:30   So how do we get from running Lumi, the company that you founded, to your work now at Obsidian's?

00:06:38   Like what was the path for you that led to this?

00:06:41   So Lumi, I ran with my co-founder Jesse for almost 13 years because it was like two startups, one after the other.

00:06:50   And we were deep working in this supply chain world.

00:06:55   And eventually around 2022, I think, the company was acquired by this company, Narvar, which is like, you've probably never heard of it.

00:07:04   But if you get packages in the mail, it's kind of like Shopify for Gap and Nike kind of like scale companies that do a lot of business internationally.

00:07:13   So it made a lot of sense for us to join with them.

00:07:17   And after a couple of years there, I was kind of itching to do something new and I was thinking a lot about what I would do next.

00:07:24   But I was having so much fun using Obsidian as part of my process to think about what I would want to do, come up with ideas about what would be next.

00:07:33   And I just realized that I couldn't think of anything that I was having more fun doing than using Obsidian.

00:07:39   And by that point, I had expanded into a lot more community projects, various plugins and tools kind of outside of the theme that we talked about.

00:07:49   And I had gotten to know Shida and Erica, who are the founders of Obsidian.

00:07:54   They were just two people.

00:07:55   You know, they're a couple based in Toronto.

00:07:58   They made this app is very community driven project and people got to know me from the community using the themes and plugins.

00:08:06   And they came to me asking for advice about like, what was it like running Lumi?

00:08:12   Lumi was a VC backed company.

00:08:14   So I knew a lot about investment going down that path.

00:08:17   And I was really impressed with how, you know, rational and long term thinking they were.

00:08:24   The fact that they didn't want to go down the VC path, which like I know very well, all of the downsides that that comes with.

00:08:32   And I think for a note taking app, it's much better to stay user funded.

00:08:36   And what would it be like to grow the team?

00:08:38   How would we approach expanding the capabilities of the tool while retaining some of its like qualities that we really love?

00:08:46   So at some point, I actually pitched to them to join the team.

00:08:50   We had been working together on the big 1.0 redesign.

00:08:55   So some of the ideas that I had created in minimal theme ended up becoming part of the 1.0 version of Obsidian.

00:09:03   And since that went well, I basically asked them if I could join.

00:09:08   And we started working together much more closely.

00:09:12   By that point, we were like four employees total with a couple, you know, community members helping out.

00:09:19   So that was the beginning of 2023.

00:09:21   So it was a very gradual and organic process because I was part of the community for like two and a half years before, three years before that even happened.

00:09:30   I think there's something so I can't think of a better word than sweet about the idea of like just really enjoying something so much that you can end up just going to work there.

00:09:43   Like there's just something so nice about that.

00:09:46   Yeah, I think that is something more people should consider.

00:09:49   In the case of me, Shida and Erica, it worked out extremely well because I think we were very lucky.

00:09:54   I don't really know if it's like super easy to replicate that path.

00:09:58   No.

00:09:59   But in that case, I think all of our skills are very complementary.

00:10:02   Shida is just like an incredible, incredible engineer.

00:10:05   Erica as well.

00:10:06   But she, I think, really brought a lot of the product and community driven nature of what Obsidian is.

00:10:13   And I think I bring something around design and communication and also just all my years of experience running startups.

00:10:20   Like I'm wearing 12 different hats doing legal and accounting and all these other things on the side too.

00:10:26   So I want to dig into some of that in a little bit.

00:10:28   But first, I want to talk a bit about your workflow.

00:10:31   It's like how you use Obsidian in your life and your work.

00:10:34   And you have a great article that kind of sums up some of this stuff that I'll make sure is in our show notes.

00:10:39   And in it, you say, I use Obsidian to think, to take notes, write essays and publish this site.

00:10:44   So you're talking about your own personal blog.

00:10:46   And you've mentioned Obsidian being a note-taking app and its fundamental level allows you to really easily link pieces of information together to create this kind of,

00:10:55   I mean, a lot of people call it like a second brain, that idea, right?

00:10:57   Like all of your thoughts are there and they're connected.

00:10:59   But with the different types of notes that you're taking, do you treat them differently?

00:11:04   Like is there different kind of rules and formatting that you might apply to, say, a note compared to an essay?

00:11:11   Somewhat, yes.

00:11:13   I think my method of note-taking is based around the knowledge that I'm going to be lazy or busy or something's going to be happening at the time

00:11:24   where I need to take a note often.

00:11:25   So I'm very much used to the idea of like chaos and laziness is the norm in terms of how I'm taking notes.

00:11:35   I might be, you know, walking down the street.

00:11:37   Suddenly an idea comes up.

00:11:38   I need to put something in there quickly.

00:11:40   So there's a whole kind of category of notes that is just like the stuff.

00:11:45   And it could come at any point.

00:11:47   I've been working a lot on this tool, Obsidian Web Clipper, that lets you grab things from the web.

00:11:53   And so maybe something catches my eye and I want to quickly save that from a website, for example.

00:11:59   So to me, all of those things kind of go into the mess of notes.

00:12:05   And I really only use a couple folders to stay organized.

00:12:09   I have clippings for those things that come from the web that I didn't create.

00:12:13   I have references, which are things like books or people, any proper noun that exists in Wikipedia, like something that exists in the world that's not one of my ideas.

00:12:24   I have essays that I'm writing or articles that I'm publishing that have a name.

00:12:31   And that name will evolve over time as I'm writing the piece.

00:12:34   I have these quick notes that I'm writing throughout the day.

00:12:37   And then I have a method of kind of progressive summarization of my diaries.

00:12:42   So, you know, if I'm journaling every day or every couple of days, then once a week, once a month, I'll try to summarize what happened in that period of time.

00:12:51   Then every six months, I try to summarize what happened every month.

00:12:55   And every year, I try to summarize everything that happened every year, using each kind of previous level of kind of zoom to inform the high levels.

00:13:03   So all of this is inherently messy.

00:13:07   There's just like stuff coming in from all different places.

00:13:10   And what really helps me is one, the concept of links, because I can always trace back where was the starting point of when I had the original idea.

00:13:19   I can always click back and back and back to where did that start?

00:13:22   I am linking a lot intentionally.

00:13:25   And the other thing that I really enjoy about Obsidian Markdown is what's called front matter.

00:13:31   So there's this structured content that you can put at the top of a Markdown note that gets formatted in a nice way that allows you to enter structured information like dates or a true or false or a status drop down or a link to another note.

00:13:47   And that's how I start to create a little bit more structure over time.

00:13:51   So if I'm watching a movie, for example, and I want to store a rating about that movie, I would store that in the front matter, the properties of the note.

00:14:01   And using these different tools, you know, links, front matter, I can find my way back to anything very easily without having to have too much overhead at the moment where the idea comes to me.

00:14:15   And then there's notes that are one sentence long, just the name of an idea, and that's it.

00:14:22   And then there's notes that are, you know, tens of thousands of words where I'm, you know, expanding on an idea or linking a bunch of things together to try and come up with a broader narrative.

00:14:32   Or I'm answering this questionnaire that I do for myself every year to kind of reflect on the year.

00:14:38   Those can become very long notes.

00:14:40   So it has no impact on the length or complexity of what's in there.

00:14:43   It's just my method for making sense of all the noise that happens every day.

00:14:48   Every time I hear people talk about how they use Obsidian, I'm always fascinated by it because I think that it is a method of organization that just doesn't come naturally to me.

00:14:58   What it essentially sounds like is you have an operating system for text, right?

00:15:02   It's just like everything goes in there.

00:15:04   Where my kind of writing, like I like to have things in buckets that are a bit more distinct because that helps me kind of, oh, I'm looking at this app now, this UI.

00:15:13   It's this kind of writing or it's this kind of note taking.

00:15:16   And I find it so fascinating to hear, like, you will make notes that are a word long or notes that are 10,000 words long.

00:15:24   And that's just all in the one system.

00:15:27   It's just like an incredibly interesting way of organizing that sometimes I can, like, look at and be like, that's amazing.

00:15:33   But I just don't know how I would even start.

00:15:35   I think you have to let go of folders and tags as a structure in order to accept this method because some people really like folders.

00:15:44   And by the way, when we get to talking about how Obsidian as a team works, we do use folders in that case because from a collaboration standpoint, it does make sense to be able to, like, go top down and see, okay, you know,

00:15:57   we want to talk about projects, we want to talk about projects, it makes sense to have a projects folder.

00:16:00   But when it comes to my own personal process, I don't know ahead of time the importance that any given concept is going to have to me later down the road.

00:16:11   And that's why I think that the Wikipedia analogy is the best.

00:16:15   Like, if you browse Wikipedia, there's entries that are a million words long and entries that are just stubs.

00:16:22   And so what ends up creating the structure is one, the links themselves, and two, for me at least, it's the front matter.

00:16:30   And last year, we launched this thing called Obsidian Bases, which allows you to use metadata from your markdown files to visualize them at a higher level.

00:16:42   So you could look at it as a table or as a gallery or people are making Kanbans.

00:16:48   There's all kinds of different visualizations, but essentially, it's taking that metadata or note level information and zooming out.

00:16:56   So for example, for me, it might be all of my journal entries, or it might be all of the movies I've seen or the places I've been to recently.

00:17:04   Even though there's no folders, I can still zoom out by creating these sort of essentially just like filtered views of content that have some specific metadata associated with it.

00:17:17   What is your basic rule for something should be a linked piece of text?

00:17:21   When you're writing a sentence, like what is the instinct that a word that you're typing or a phrase that you're typing is something that should be linked?

00:17:30   Because the way, if I'm understanding the way to try and explain how Obsidian works, you're essentially creating these links that aren't necessarily linking to something else directly.

00:17:39   It's not like a hyperlink, but just by creating a linked piece of text, anytime that you reference it again in the future, you can always kind of find every place that you've selected, essentially like a wiki, as you say.

00:17:52   Yes, and you don't even need to create that note yet.

00:17:57   So if I was typing, today I spoke to Mike, I could make Mike a link.

00:18:03   And making a link is very easy in Obsidian.

00:18:05   All you do is type this double brackets notation around a piece of text.

00:18:11   And that link can exist as unresolved is the term, but it just exists.

00:18:16   It's known within the Obsidian app that there's a reference that doesn't yet have a note.

00:18:22   And so to answer your question, pretty much any time I'm mentioning the name of a place, a person, a book, a concept, I will make it a link the first time in the note.

00:18:35   I don't make it a link for every single mention of that thing within the note, mostly because it's just too much work and not necessary.

00:18:43   There are different kinds of things that are links in my system that may not be for other people.

00:18:49   For example, I have a note for every day.

00:18:51   So every day has a timestamp and there I can view anything that happened that day or was linked to to that day, which is really fun because sometimes things, dates could be in the future or in the past.

00:19:05   And then connections can exist later on, like 10 years later, I find out, oh, something happened on that same day that I did that other thing.

00:19:13   And you just kind of find those connections way later.

00:19:18   Another concept that I really like is evergreen notes.

00:19:22   It's a little tricky to explain, but it's basically a concept.

00:19:25   You can think of an idiom as an evergreen note.

00:19:29   It's just an idea that you can formalize into a sentence and then that sentence becomes a link.

00:19:36   You know, on the homepage of Obsidian, we show an example of that, which is, I think, therefore I am from Descartes.

00:19:43   So that idea, if it has meaning to you, can become almost like a meme inside of your own knowledge base where I keep coming back to this idea.

00:19:54   It's a little bit of a shorthand or shortcut, just like an inside joke might be to kind of reference a whole world of ideas that are hidden behind that link.

00:20:13   But because you have that shorthand for yourself, you can quickly form bigger ideas out of these kind of Lego blocks.

00:20:21   I think if you have these kind of phrases that summarize a feeling that you're having, it would otherwise be incredibly difficult to search for every note in which you've had that feeling.

00:20:31   But by linking them together with these links, you're creating a manual linkage that would otherwise be lost, I guess.

00:20:38   Yeah.

00:20:38   And, you know, in the context of the Obsidian team, for example, a link might be, we call them directives inside of Obsidian, but it's a general thing that we're working towards.

00:20:51   It's like a meta project.

00:20:52   So for example, localization, we're trying to make sure that Obsidian is easy to use in every single language around the world, which is quite difficult.

00:21:00   And so anytime we're inside of our shared vault for the Obsidian team, and we're talking about something that could touch localization, we will link back to that evergreen note that has to do with, you know, Obsidian should try to work in every language.

00:21:16   This episode of Cortex is brought to you by FitBond.

00:21:20   If you're looking to make changes to your personal fitness, getting started can feel like an uphill battle.

00:21:25   That's why I'm pleased to let you know FitBond is an easy, approachable, and affordable way to build a fitness plan that is just for you, because everybody has their own path to personal fitness.

00:21:35   And it's about understanding what is going on with you, your body, your experience, your environment, your goals.

00:21:42   This is what makes a difference.

00:21:44   FitBond takes this information, stores it in your FitBond gym profile, and is tracking your muscle recovery.

00:21:50   To make sure that you're avoiding burnout and keeping up your momentum, while also designing a well-balanced workout routine for you.

00:21:56   This means you're not going to get bored, but it's also going to be mixing up your workouts, which means that your muscles are going to work together.

00:22:03   You're not going to overwork or underwork anything.

00:22:05   This will negatively impact your results.

00:22:07   So FitBond tracks your muscle recovery to make sure it's building the best possible workouts for you.

00:22:12   It combines AI and exercise science.

00:22:14   FitBond have analyzed billions of data points that are fine-tuned by certified personal trainers.

00:22:19   And they have more than 1,000 demonstration videos in the app, so you're going to make sure you're learning new movements the right way.

00:22:25   That is my favorite thing about FitBond, because it's always giving me new exercises, which is great because I don't get bored.

00:22:30   It mixes things up, keeps me engaged, but it also shows me how to perform every workout, and that is awesome.

00:22:37   The app is super easy to use.

00:22:39   You can stay informed of their progress tracking charts, weekly reports, and sharing cards, so you can keep track of your achievements, your personal bests, and share them with your friends and family as well.

00:22:47   FitBond also integrates with your Apple Watch, your Wear OS, SmartWatch, and apps like Strava, Fitbit, and Apple Health.

00:22:54   Personalized training of this quality can be expensive. FitBond is just $15.99 a month, or $95.99 a year.

00:22:59   But you can get 25% of your membership by signing up at fitbod.me.cortex.

00:23:04   So go now and get your customized fitness plan at fitbod.me.cortex.

00:23:09   One last time, fitbod.me.cortex for 25% off your membership.

00:23:13   Our thanks to FitBond for their support of this show and Relay.

00:23:17   So obviously Obsidian is a huge part of your life, but I'm assuming you use other software.

00:23:22   So what are you using for task management?

00:23:26   Say for your personal tasks.

00:23:28   Maybe you use Obsidian for the team tasks, but what are you using for yourself?

00:23:31   Yeah, I do use Obsidian, but my task management is very lightweight.

00:23:35   I just, I have an article.

00:23:38   It's sort of trying to be provocative because I feel like people really overcomplicate task management.

00:23:44   But I just create a note once a week that says a list of things that I want to do.

00:23:50   And I come back to it throughout the week.

00:23:53   I'll check things off that list.

00:23:54   And then the next week I start a new weekly note and I write a new to-do list.

00:24:00   And I try not to look at the past week's to-do list.

00:24:04   I just try to write it from scratch.

00:24:05   And usually that automatically filters things up to the top that needs to be top of mind.

00:24:12   And then I might look at the previous week and roll over anything that didn't get done.

00:24:17   And at that moment, I might also toss out some things that I thought I needed to get done,

00:24:22   but it turns out it's not really that important.

00:24:24   So it's very simple.

00:24:25   When you mentioned the memory thing, I was really worried that you were going to say that

00:24:29   if you didn't do it, you just left it.

00:24:31   Like, you know, just like, ah, it's gone to time.

00:24:34   I do do that actually, kind of.

00:24:36   I mean, there are these tasks that kind of get rolled over week after week after week.

00:24:40   And I look at them and I'm like, and eventually sometimes they just disappear.

00:24:44   It's like, it's like emails.

00:24:47   Like sometimes it just, you know, you got to archive all.

00:24:49   Have you ever done this on pen and paper?

00:24:51   I used to, but now I think I just,

00:24:55   I have too many notebooks.

00:24:56   I have so many notebooks everywhere.

00:24:58   That is the notebook problem.

00:25:00   Yeah.

00:25:01   I do still use notebooks sometimes, mostly for sketching or if I'm outside,

00:25:06   like if I'm at a restaurant or if I'm hanging out with a friend,

00:25:09   I don't really like to take my phone out.

00:25:11   So I still use that and I might jot something down quickly,

00:25:14   but it always ends up going back into Obsidian now.

00:25:17   The way that you manage your tasks in Obsidian,

00:25:21   it seems like the closest to a pen and paper system that I've heard of,

00:25:25   is like actually writing them down fresh every time.

00:25:28   I think it's what a lot of people like about using pen and paper for task management,

00:25:34   is that practice of being forced to review them again.

00:25:37   Yes, I totally agree with that.

00:25:40   Yeah.

00:25:40   I think that's why email is so hard to manage because it's like everyone's putting a task into your inbox and you have to clean that up somehow.

00:25:51   Whereas if you go down this path, which is not for everyone,

00:25:56   but the nice thing is it does help you focus.

00:25:59   You mentioned email.

00:26:00   What is your relationship to email?

00:26:02   I basically don't use email and don't use a calendar.

00:26:06   Wow.

00:26:06   I have an email inbox.

00:26:08   I will check it every day.

00:26:10   But most of the stuff that goes in there is a bunch of notifications from various other systems.

00:26:16   So I'm pretty aggressive about clearing it out once every couple months.

00:26:23   But there will be multiple days where I go without checking email.

00:26:26   For me, email is mainly a method of communicating with my grandparents and my parents.

00:26:32   I mean, it's also why Obsidian, the team and the structure is designed the way it is.

00:26:38   We don't have meetings.

00:26:39   We don't have a sales team.

00:26:40   We don't have anything that would require me to have emails.

00:26:44   It's sort of like it's designed around the idea of we don't have to have emails.

00:26:49   So the work is it's kind of the opposite of how the majority of people might work.

00:26:54   For us, the team and the company is constructed around making it possible that we don't need emails and calendars.

00:27:02   Okay, I couldn't imagine living without a calendar because I know I would not do all of the things that I do because I collaborate with a lot of people.

00:27:09   And it tends to be through structured times.

00:27:12   So when you say no calendar, do you really mean you never use a calendar?

00:27:17   For me, calendar is just to remember that I have a dentist appointment.

00:27:21   We have no scheduled meetings at Obsidian.

00:27:24   So we have one meeting a year that is an offsite.

00:27:27   We usually go somewhere in the world, you know, for a week with the team.

00:27:32   So I know that I'm going to have that at some point.

00:27:34   I know that I'm going to have a couple of doctor's appointments.

00:27:36   And you hope you remember the day so you're going to arrive.

00:27:38   No, no, no.

00:27:39   Technically, a calendar does exist.

00:27:42   And we will put events in there for the team.

00:27:45   And I have some events here and there.

00:27:47   Like if I have a dentist appointment, I will put it into like the calendar app.

00:27:51   I am much more a user of the reminders app, the Apple reminders app.

00:27:56   So there are things like an electrician is going to come on Tuesday to fix this problem that we have.

00:28:04   And I will typically basically just use Siri to create a reminder that that's going to happen.

00:28:12   Okay.

00:28:12   But I put the reminder at the moment that I will need to remember that information.

00:28:17   Right.

00:28:17   So maybe I need to know like the evening before or the morning of that that thing is happening.

00:28:23   And that thing might exist in the calendar if I remember to do it, although most of the time it doesn't.

00:28:29   I mean, for you, your work, it would be kind of impossible to do without a calendar because you're scheduling things with everyone all the time.

00:28:35   Yeah.

00:28:36   But we don't need to do that.

00:28:37   So I don't currently have that need.

00:28:40   By the way, I worked for 15 years with a calendar all the time.

00:28:44   So this is a conscious choice of me removing calendar.

00:28:48   But it's not like it magically happened that way.

00:28:50   It's an intentional choice.

00:28:52   I could imagine if I had a work transition, I would not want to use a calendar anymore either because I feel chained to my calendar.

00:29:00   Yeah, when I was back in the Lumi days, I would probably have back-to-back meetings all day long.

00:29:06   I would have 8 to 10 meetings a day.

00:29:08   And so I was, you know, in the calendar hellscape for many years.

00:29:12   So I know what that is like.

00:29:14   And that's why I designed my entire life around not having to have a calendar.

00:29:18   But I get it.

00:29:20   It's like there is a calendar, but it is not a part of your structure in life.

00:29:27   It's not a part of your routine to check it because the things that you're doing won't be there most of the time.

00:29:33   Yeah.

00:29:33   So I did put a calendar and a reminder that we were going to have a podcast today.

00:29:37   And that is helpful because the downside is it does occur once in a while that I just totally forget something is happening.

00:29:45   It happens less these days because I think I've narrowed down my reminder system pretty well.

00:29:51   Yeah.

00:29:51   But it does happen once in a while that I forget.

00:29:55   But what I like is just looking at my calendar and just knowing I have all these chunks of uninterrupted time where I can do whatever I want in those blocks.

00:30:03   And that to me is freedom.

00:30:05   So let's talk about the team at Obsidian and how you all work together.

00:30:10   You've mentioned a couple of numbers throughout history.

00:30:13   Right now, how many people work at Obsidian, including you?

00:30:17   Full time, I think seven people.

00:30:20   We have three engineers, including Shida, the CTO.

00:30:24   We have Erica, myself.

00:30:26   And then we have Rebecca, who's our customer service person.

00:30:30   And we have Joe, who is kind of a community lead who works kind of managing the plugin ecosystem.

00:30:38   And then there's a handful of other people who are in the orbit who either work part time or they're like moderators who are volunteers.

00:30:49   So there's kind of like this extended network of people that we collaborate with frequently, but that are not full time team members.

00:30:57   Yeah.

00:30:57   And I would assume for moderation, you know, you have very active communities like Discord and forums and such that require that kind of help, I guess.

00:31:07   But what really helps everyone, whatever your relationship to the company is, is we have this manifesto on our about page that's very short that explains the core principles of Obsidian.

00:31:20   And that's the main alignment that everyone is referencing.

00:31:23   And so the nice thing about that is the reason you use Obsidian is because you like those ideas.

00:31:30   And so it kind of flows naturally what we should do, what the community should do from looking at that set of principles.

00:31:38   I get the sense that you like the team to be small.

00:31:43   Again, this is in a way a reaction to how I worked before because I've ran a startup that had 45 people.

00:31:50   I worked in a company that had a few hundred people when we were acquired.

00:31:54   And I never had more fun than when we were 10 or less.

00:31:59   And I think the reason why is because personally, I really enjoy being a contributor.

00:32:04   I like designing.

00:32:06   I like coding.

00:32:07   I like writing.

00:32:07   I want to do that most of my time.

00:32:10   And so I think maybe you could stretch it to 15 or 20.

00:32:14   Some companies I know like Valve experimented for a while with, you know, hundreds of people who are all no managers type of environment.

00:32:22   There's a lot of experiments that have been done around that concept.

00:32:25   But I think just having a small team allows everyone to kind of keep track of what everyone else is doing and know each other pretty well and each be somewhat of a generalist, but also have their own core skills that they're really great at.

00:32:40   And again, to not have to manage people and have suddenly one-on-ones and all these different things that you have to do to kind of keep the context alive across everyone.

00:32:52   And so, yes, this comes with, okay, maybe Obsidian is not going to be as big as it could be if we had hundreds of people on staff or something like that.

00:33:01   But I don't mind that constraint.

00:33:03   Obviously, it feels right for you and it feels right for the team.

00:33:06   Do you think it's right for the product?

00:33:08   I make Obsidian because I want to use it.

00:33:11   And I think everyone on the team feels that way.

00:33:14   We want to keep improving Obsidian for ourselves.

00:33:17   And, you know, maybe this comes with frustration for some of our users who wished that we would prioritize a feature, perhaps, that matters to them.

00:33:25   But no one on our team actually cares about that particular thing because in our personal workflows with Obsidian, it doesn't actually matter as much.

00:33:34   I think the downside of that is, you know, what I just described, we don't have as much breadth of time and specialists working on all these different areas at the same time.

00:33:46   On the other hand, I think the good thing is it keeps us more focused.

00:33:51   We kind of always know what our top priorities are.

00:33:54   We're always continuing to use everything that we put out.

00:33:57   So, like, the capabilities that we add to the product, there's not these major areas of unknown capabilities that nobody's using.

00:34:05   We're actively using it and caring.

00:34:08   I don't want Obsidian to be suddenly filled with, like, these magical AI buttons everywhere because I'm using the app all the time and I don't want that, you know.

00:34:16   So, the escape hatch is that we have this amazing plugin system where anyone can build whatever they want on top of the product.

00:34:27   So, that kind of allows us to stay more constrained and focused on what we think Obsidian should be and how it should work.

00:34:35   So, it has its pros and cons.

00:34:37   But for me, the main priority is, is it fun to work on every day?

00:34:41   Am I going to leap out of bed and want to work on it?

00:34:43   And if I was running a company with 200 people doing all random different things, I don't think that would be true.

00:34:49   Because I can imagine the app is so important to you, obviously, right?

00:34:53   It helps if it's important to everybody else.

00:34:56   And I feel like the bigger you make your team, the less likely that will be.

00:35:01   There's just a diffusion of responsibility.

00:35:03   There's more different people who care about different things.

00:35:07   I really care that I can know every single screen that exists in the app, which there are a lot of.

00:35:12   And I like the responsibility that it puts me under to have that knowledge, that deep knowledge of the product.

00:35:18   Do you make an effort to familiarize yourself with everything?

00:35:22   Because I can't imagine every feature is something that you want yourself.

00:35:28   Like if something's being added to the app that is not necessarily what you would have in your system, that you really try to understand it?

00:35:36   Yes.

00:35:37   I try to figure out how I would make that capability useful to me.

00:35:42   And then that makes it a lot easier to work on.

00:35:45   The Web Clipper that I mentioned a while ago was a good example of that.

00:35:49   For many years, we didn't have that capability.

00:35:52   And it's something that people who are coming from Evernote or some other tools really loved, being able to save things from different places.

00:35:59   And so there was enough momentum from the community that people kept asking about it and saying,

00:36:05   Oh, you know, I wish I could switch to Obsidian, but this one feature is missing.

00:36:10   And I didn't care as much about it personally because I try not to bring too many external things into my personal vault.

00:36:19   I'm very parsimonious about like the things that I will bring in.

00:36:23   But after people asked enough, we like to call it complaint-driven development.

00:36:28   It's just like if enough people complain, we'll start to think about it.

00:36:31   Because you assume at that point, it's like, I can't see this, but clearly everybody else can.

00:36:37   Is it me that's the problem?

00:36:39   Yeah.

00:36:40   So I thought, okay, is it me that's the problem?

00:36:42   And I started thinking about, okay, what would I want out of a Web Clipper?

00:36:46   And so I started building it, and I think it turned out very different from any other tool that is like it.

00:36:54   And now it's evolved quite a bit over the past year where there's some really great capabilities that we just launched,

00:37:00   like being able to just read websites.

00:37:02   And a lot of browsers have this like reader type of functionality that will strip away all the like navigation and junk that's on a page.

00:37:10   But how do you actually make that really good?

00:37:12   And how does it work on YouTube or like multimedia sites?

00:37:15   So I've been working on that.

00:37:16   But that came from me understanding what the community was asking for and trying to like reconceptualize it in a way that I can build into my personal workflow that it makes it fun for me to work on.

00:37:30   Because I just, it's hard to have the motivation to work on something if I'm not going to end up using it or caring about it.

00:37:36   This episode of Clotex is brought to you by Sentry.

00:37:40   Applications break in all kinds of ways, from crashes to slowdowns to regressions, the stuff you only see once real users hit them.

00:37:48   Sentry catches all of it.

00:37:51   That's S-E-N-T-R-Y.

00:37:52   With Sentry, you get traces, replays, errors, profiles, and the details around them like stack traces, commits, releases, and the developers who broke it.

00:38:01   All in one connected view.

00:38:03   So you're not just jumping between tools trying to figure out exactly what happened.

00:38:07   Sentry will show you how the request moved, what ran, what slowed down, and what the user saw.

00:38:12   And then SIA, Sentry's AI debugging agent, will take it from there.

00:38:16   It uses all of that Sentry context to tell you the root cause, suggest a fix, and open a pull request for you.

00:38:23   It also views your pull request for you and flags breaking changes with a fix ready.

00:38:29   I love apps that work well.

00:38:30   If I hit a bug, I kind of want it to be fixed, and I hope that the developer's going to do something about it, and quickly.

00:38:36   That's what we all want.

00:38:37   We want to get the experiences we want from the applications that we love.

00:38:41   And what I love about Sentry is it gives developers the tools to be able to tackle these things quickly.

00:38:46   And what I know from working with developers is they want to be able to deal with these things fast so they can get back to doing the new stuff that they really want to give you as a user.

00:38:55   I love this.

00:38:56   I love that Sentry provides these tools to the developers whose work I enjoy.

00:39:01   Try it out free at Sentry.io, that's S-E-N-T-R-Y.io, and tell them that we sent you.

00:39:08   They have a free dev plan, and listeners of this show can use the code CORTEX26 to get $100 in Sentry credits if you're a new user.

00:39:16   That's S-E-N-T-R-Y.io and the code CORTEX26.

00:39:20   Our thanks to Sentry for their support of this show and all of Relay.

00:39:24   All right, so you've mentioned a few times no meetings.

00:39:27   I want to understand if this really, truly means no meetings.

00:39:31   You don't have a stand-up or anything like this?

00:39:33   Right, so right now we do have a team calendar.

00:39:37   This exists, but the only thing that's on there right now is a scheduled off-site that we're going to have this summer.

00:39:45   And we have had a meeting recently, about a week ago we had a meeting.

00:39:50   Because we're working on this very complicated project that involves a lot of different people.

00:39:55   It wasn't on the schedule, it's just like, hey, I think we should talk about this because it's touching a lot of different areas and it would be good for us to kind of sync up on it.

00:40:04   Yep.

00:40:04   So we had a meeting for that.

00:40:05   But there's no regular check-ins.

00:40:08   People are welcome to meet with each other.

00:40:10   There are ad hoc things that happen.

00:40:12   It's not like meetings are disallowed.

00:40:14   It's just that the assumption is everyone is contributing to the product in some way and they should have deep chunks of uninterrupted time.

00:40:26   And we also have people who have small children or different medical challenges or any, like, people, we just want people to be able to live their life.

00:40:36   And everyone's excited enough about working on Obsidian that I'm not concerned about that falling through the cracks, you know.

00:40:42   It's a very interesting setup.

00:40:45   I mean, something like this will always come from the top down, right?

00:40:50   That if you like to have a meeting every Monday, people are going to start booking meetings with each other, right?

00:40:58   Because that becomes the culture of the organization.

00:41:01   And if you do not do that, then people naturally would not do that.

00:41:06   I think meetings are a function to force collaboration, right?

00:41:11   Like, I think that's why they exist in organizations.

00:41:13   You get people together and they talk and they understand each other and they can all go away and do their work.

00:41:18   But it's not the only way you can do it.

00:41:20   So how do you do it?

00:41:22   How do you encourage collaboration amongst your team?

00:41:26   How do you collaborate with others on your team?

00:41:29   How are you keeping track of what everybody's doing and where necessary, kind of like working together?

00:41:33   So the way the Obsidian team functions is we intentionally made certain choices and they're probably not going to work for everyone.

00:41:42   They might not be desirable for everyone.

00:41:44   There's certain things, depending on the scale at which you're operating, that it's just probably impossible to do it without meetings to some extent.

00:41:52   So this is not like a recommendation.

00:41:54   It's just a, here's how we designed it.

00:41:57   We primarily use Discord as our chat interface, kind of similar to Slack or other tools like it.

00:42:04   And throughout the day, we are constantly talking on there.

00:42:10   So we have channels that are around certain projects or certain capabilities or certain things like, you know, customer support or things like that.

00:42:20   So there are conversations that are going on throughout the day.

00:42:23   People are in widely different time zones.

00:42:26   So we try to have a workflow that is compatible with asynchronous work.

00:42:32   So there are times where two people might be collaborating very closely on some particular feature.

00:42:39   And so asynchronous conversation over chat might sort of materialize out of thin air and people kind of know where everyone's time zones are.

00:42:48   And so that happens naturally.

00:42:49   But then there's also a lot of asynchronous discussion that happens where throughout the day, people are posting status updates about what they're doing.

00:42:58   It's just kind of broadcasting like a heartbeat of what's going on.

00:43:01   So, you know, I'm working on this thing.

00:43:04   I'm running into this problem.

00:43:06   It's not meant for anyone to reply to.

00:43:09   It's just to verbalize what is currently going on.

00:43:13   And sometimes it helps kind of as a rubber ducking to just be talking into the ether and just, oh, okay, now actually I have these different approaches that I'm thinking of and people might jump in with some thoughts or ideas along the way.

00:43:27   We have this concept called ramblings, which emerged sort of ad hoc naturally, which is channels that are publicly visible to the rest of the team, but that are for it's kind of like an internal Twitter.

00:43:40   You know, you can just post what's going on or anything interesting, like an article that you read recently that is relevant to a project that you're working on.

00:43:49   Or it could just be yesterday I posted, I saw this amazing rainbow when I was walking my dog and that was it, you know, especially when you're remote, just sharing what's going on in your life.

00:44:00   And people can see that no one can post to your ramblings channel, but they can reply.

00:44:05   So it's a nice way to do what you would otherwise do as like a water cooler conversation in an office.

00:44:13   So I feel like you've touched in a few places about kind of like the structure working for the business, the team size working for the business.

00:44:21   So Obsidian's business model is the app is free, essentially for people to use without limits, but you have some optional paid tools and these seem like things that would bear some cost to them.

00:44:38   And then you obviously make money from that as a business as well.

00:44:42   Do you feel like this business model works for Obsidian?

00:44:46   Yeah.

00:44:47   Yeah.

00:44:48   Like you said, the app is free.

00:44:49   So there's no features of the app that are blocked.

00:44:53   There's no like gated capabilities.

00:44:55   There's no like maximum number of notes or something like that.

00:44:58   You don't even have to sign up for an account.

00:45:01   You can just download the app and start using it.

00:45:03   What is paid are our services.

00:45:05   So we have sync service and a publish service.

00:45:08   And those are optional.

00:45:10   You can also just use Git or Dropbox or any number of different tools out there.

00:45:15   So our goal with our syncing and publishing services are to make them the best that we can for the use case of Obsidian.

00:45:25   And we charge for those.

00:45:27   And that's been working great.

00:45:29   But also because we're small and we want to be, you know, stay the size that we are and we don't have any investors breathing down our neck.

00:45:37   We're not trying to become the next Microsoft or something.

00:45:40   We're not trying to take over the world.

00:45:42   So it works.

00:45:43   You know, it's profitable.

00:45:44   And as long as we are able to keep the lights on and keep the team paid and keep having a chance to work on this, I don't see that stopping anytime soon.

00:45:55   So, yeah, we're happy with it.

00:45:57   Do you ever come into conflict with this business model and a software business?

00:46:02   Like are there things that are inherently more difficult for you to do because you're not maximizing your revenue potential?

00:46:11   Yeah, well, a good example of that is previously we had one of our revenue streams was the commercial license.

00:46:17   So when Shida and Erica launched Obsidian, this was the only revenue stream that existed because sync and publish didn't exist.

00:46:25   But there was terms in our user license that said, if you're using this for personal use, it's free.

00:46:31   But if you're using it for use within a company of two or more people, then you have to pay for the commercial license, which is $50 a year.

00:46:39   So that was true for about four years of Obsidian's life.

00:46:44   And then last year, we actually made the commercial license optional.

00:46:47   So we said anyone can just use it.

00:46:50   If you want to buy a commercial license, you can to support the project, but you don't have to.

00:46:55   And so that was a shift that came from a necessity because essentially what we realized was since you can download the app and don't have to sign in, the commercial license was purely a honor system thing.

00:47:12   And it was inspired somewhat by another app called Sublime Text, which I love that has a very similar model.

00:47:17   But the problem is that there was no way to enforce that.

00:47:20   And we knew that there were companies with thousands of employees using Obsidian that were just not paying and kind of just skirting the terms of service.

00:47:28   And so we decided to just make it optional because it was more reflective of the reality.

00:47:36   And so we left probably millions of dollars on the table making that choice.

00:47:40   But what it enabled was that there were all these people who wanted to use Obsidian for their work notes, but were afraid of breaking the license.

00:47:51   So we just said, let's just remove that constraint.

00:47:55   Anyone who wants to use it is free to use it.

00:47:58   And the hope just being that because people will then use it, they will share with their friends or coworkers and eventually that transfers into sync or publish revenue.

00:48:08   But we have no way of measuring this because we had no way of measuring how many people were like not paying us for the commercial license before.

00:48:17   We don't really have a way of measuring how many active users Obsidian has because we don't have analytics in the app.

00:48:23   And if they're not using one of the services, you just don't know that they exist, right?

00:48:27   Exactly.

00:48:28   I mean, we can kind of guess based on certain numbers like GitHub downloads and the app stores have download numbers.

00:48:35   So we can have a general idea of how many people overall downloaded the app.

00:48:39   Yeah.

00:48:39   You know, I think at other companies, this would make everyone extremely uncomfortable.

00:48:53   But because we don't have investors and our priority is creating an app that we're proud of and that our lifestyle building this app is fun and enjoyable for everyone, we're okay with the tradeoffs of like not knowing what that actually means for potential lost revenue.

00:49:10   And I imagine as well in removing the kind of enterprise agreement, you don't have to live up to all of the enterprise support stuff that would potentially come because I imagine that's more complicated than user support.

00:49:25   We ran into that a lot with big companies.

00:49:27   I mean, basically, 10s of 1000s of very large companies are using Obsidian and some of them have subsidiaries in dozens or 100 countries.

00:49:37   And what we were finding back before the commercial license change was there were teams that were individually buying commercial licenses from all these different countries.

00:49:48   And basically, this kind of parent company was like, how many users do we have?

00:49:53   Or how many licenses do we have?

00:49:55   And we're like, well, we have to go and like tally these numbers from every different subsidiary of the company.

00:50:01   And it was literally like over 100 subsidiaries of the company, because they just wanted to consolidate the billing of their commercial licenses.

00:50:10   But then the person who was doing that ended up leaving the company and that whole project fell apart.

00:50:16   So like not having to deal with all of that is quite nice.

00:50:19   This episode of Cortex is brought to you by our friends at Squarespace, the all-in-one website.

00:50:24   One website platform that is designed to help you stand out and succeed online.

00:50:29   Whether you're just getting started or scaling a business, Squarespace will give you everything that you need to claim your domain, showcase your offerings of a professional website, grow your brand and get paid all in one place.

00:50:39   If you're running a small business, you want to be able to do it online and you want to be able to be online and you want to do it quickly.

00:50:43   What I love about Squarespace is they give you all of the tools that you need to be able to show your business to the world.

00:50:48   For example, you can get discovered fast with integrated SEO tools, because why build this beautiful website for nobody ever to find it?

00:51:03   Every Squarespace website is optimized to be indexed with method descriptions, an auto-generated sitemap and more, so people find your site through search engine results.

00:51:12   And you can also offer your services and get paid with Squarespace.

00:51:16   You can offer consultations, events, experiences, and so much more.

00:51:19   You can showcase your offerings with a customizable website that is designed to attract clients and grow your business.

00:51:25   Think about built-in appointment scheduling, email marketing tools, and so much more.

00:51:29   You can also keep everything cohesive with on-brand invoices and get paid easily with online payments.

00:51:34   But don't just take my word for it.

00:51:36   Go and build your website today by going to squarespace.com slash cortex.

00:51:41   You'll get a free trial to do so.

00:51:43   You can build your whole site, and when you're ready to launch, use the offer code cortex,

00:51:47   because you will save 10% off your first purchase of a website or domain.

00:51:51   That is squarespace.com slash cortex and the offer code cortex to get you 10% off your first purchase and show your support for the show.

00:51:59   Our thanks to Squarespace for the continued support of Cortex and all of Relay.

00:52:03   So I get the sense that growth is not something that you're focused on chasing.

00:52:09   Growth is a weird term because I think what you probably mean is like revenue growth or user growth or something like that.

00:52:15   Anything, yeah.

00:52:15   Well, right now we're seeing from everything that we can see with the few data points that we have,

00:52:21   the adoption of Obsidian as an app is just like really skyrocketing right now.

00:52:26   And it's hard to put exact numbers around that, but I hope that the outcome of that is people have better ideas.

00:52:34   They have more nuanced views on things that they are able to do things that they weren't able to do otherwise.

00:52:40   You know, for everyone, it might be a little bit different what Obsidian brings to them.

00:52:45   So it does give us a fuzzy feeling to know that lots of people are using it.

00:52:50   But I do think that even if no one was using Obsidian, I would still be working on it for myself because that's what I did for many years before Obsidian existed.

00:53:00   And maybe it wouldn't be able to pay for my mortgage, but it would be able to give me some personal fulfillment.

00:53:08   Do you know why adoption is skyrocketing right now?

00:53:12   Or do you have a sense?

00:53:13   There's a few big things.

00:53:15   One is AI is, you know, taking over everything these days.

00:53:19   And people are switching to Obsidian both because they want to use AI and because they don't want AI, which is kind of confusing.

00:53:28   Obsidian has no AI features in it.

00:53:32   So a lot of people who are kind of trapped in other ecosystems are finding themselves like swamped in all this AI junk that is like polluting all of the interfaces.

00:53:42   And they do not want AI in their note taking.

00:53:45   And so they're switching to Obsidian because Obsidian is private.

00:53:48   It has no built-in AI features.

00:53:51   You can just use it and not be bothered by all of that.

00:53:54   And I think it's becoming quite popular for that reason because of AI infiltrating everything in other tools.

00:54:01   At the same time, because Obsidian is just based on plain markdown files and all it really does is it looks at a folder on your computer and turns all of those plain text files into something more like easy to read and easy to write.

00:54:19   And AI is extremely good at dealing with plain text files.

00:54:23   People are finding that they're connecting existing tools, whether it's like Cloud Code or Codex, Cowork, or there's all these different tools that exist that are generating a ton of markdown files.

00:54:35   They're using Obsidian as a way to view and interact with their agents, their AI-created workspaces, or to just collaborate with LLMs.

00:54:47   So it's quite confusing that both of those things are happening at the same time.

00:54:51   But it's good in the sense that we, because of plugins, but also just the file system, the fact that Obsidian is reading files on your disk, it opens the door for either of those things to be true at the same time.

00:55:03   It's like, what a fascinating, lucky turn of events, right?

00:55:08   It's just like the principles that you have as a business have put you in a place that other companies are seriously chasing, which is like people are using Obsidian to work with their AI of choice.

00:55:21   Yeah.

00:55:23   Because it's all easy to read for the LLM.

00:55:25   There's this great essay, Always Bet on Text, that you should link to.

00:55:30   And File Over App is one that I wrote that kind of resonates with that.

00:55:33   As well, where the reason that we went down the path of plain text is because we're honest that no app really lives forever.

00:55:44   Apps have a lifespan because operating systems change, user behavior changes, like the input devices change.

00:55:52   And I'm certain that Obsidian is not going to live forever, like fundamentally, it's going to eventually run out of steam or like things will change, I'm pretty sure that I will outlive Obsidian in a certain way.

00:56:05   And I want my knowledge and my notes and my notes and my journal and all the stuff that I've created over my life in digital form to stick with me.

00:56:15   And so we always wanted to think about how we can make that information durable.

00:56:21   And the simplest way that we could think of is let's go back to the beginning of computers.

00:56:26   And what has existed since the 60s are plain text files.

00:56:30   So if we want to create a system that is likely to still work in 100 years, or if my grandchildren want to read my notes for some reason, or some future aliens discover a hard drive, like if we can make our information and digital data kind of work with some of the oldest systems, we hopefully will ensure that they will keep working in the future.

00:56:53   And that was the fundamental reason to choose plain text files on your hard drive, it turns out, and I don't know if it's luck or karma, or it's just logical, that LLMs at the end of the day, they do very complex things, but they are just processing individual characters.

00:57:12   And so the more efficiently you can do that, the better it is for the computer and the better it is for the human.

00:57:19   So it kind of naturally makes sense that that's where it would end up.

00:57:23   The thing that was unnatural was this past 15-20 years where everyone was uploading their data into somebody else's server in a mostly unencrypted way, and you would be renting access to your data from these providers who may go under, may have data leaks, may have some number of different problems, may decide to extract more money from you.

00:57:48   So there's a whole host of different things that would occur, and I think the pendulum is swinging away from that.

00:57:55   How do you feel about AI?

00:57:58   I think AI is mostly going to end up being a good thing, but AI is sort of like a very vague term.

00:58:06   I don't even know if anyone would agree on what a definition of AI even means right now.

00:58:10   Yeah.

00:58:11   There's an essay I wrote a few years ago called Don't Delegate Understanding, and it's about the idea that if you start to take advice for granted or suggestions or recommendations for granted, as opposed to understanding them from scratch, you're sort of building up this shaky foundation on which you're making decisions at this higher level that are not backed by an understanding of the underlying things.

00:58:41   And I think I'm just very lucky to have been born with a lot of curiosity.

00:58:46   I get really fascinated by all kinds of things.

00:58:48   Like I mentioned, I have an appointment with an electrician coming over to fix this thing.

00:58:54   If the electrician is here, I'm absolutely going to spend time trying to learn what's going on.

00:58:59   I'm not an electrician, but I want to understand how my house works and how these outlets are connected and all this different stuff.

00:59:05   And if at every turn you kind of outsource that understanding of how things work, it's very easy to be sort of like covered in parasites that are making it seem like it's easy.

00:59:19   Essentially, whether it's through for financial reasons or control or power, extracting your time, your money, when simply understanding how the system works would give you much more control and would help you make better decisions.

00:59:36   So my biggest fear with AI and why I personally don't really use AI in my own note taking practice or we don't really use it that much for programming at Obsidian is that we want to have a really, really firm understanding of our code base, our ideas, our knowledge.

00:59:53   Where AI, I think is useful is it can lighten the load for complex tasks that otherwise would take a lot of energy.

01:00:01   So then don't delegate understanding is kind of my negative spin on AI.

01:00:06   My more positive one is another essay I wrote called Caloric Energy is Precious, which is kind of saying like we're moving into this era where in the past, before the Industrial Revolution, anytime you wanted to do something,

01:00:21   an animal or a person had to eat food and then do a thing, you know, had to pull equipment through muscles and calories.

01:00:31   And what we're doing right now is we're saying, actually, we can do the same thing for certain knowledge work.

01:00:38   Instead of using calories that require me consuming a piece of food and then doing some work with my brain, we can turn that into electrical energy to do that same work.

01:00:49   And what will you use that work for?

01:00:52   Hopefully, it's things that you have a good understanding of.

01:00:55   So an example of how we're using AI right now at Obsidian is localization.

01:01:01   Talked about that, you know, a little bit ago, but we're trying to make Obsidian work really well for all users in all languages.

01:01:08   And right now we're trying to support around 40 to 50 languages, and that includes a lot of right to left languages like Arabic and Hebrew that the entire interface is mirrored.

01:01:20   There are languages like Khmer, which is like a language that is really difficult for LLMs to deal with because every character takes up like 15 times as much space in terms of like bytes as English.

01:01:34   There's all these different challenges, and we have our help documentation, our website, our app, all of the app stores where the marketing is in different languages in different countries.

01:01:44   And we're only seven people on the team.

01:01:46   There's no way we're going to hire 40 different translators in every language to cover our help desk information, which has like the contents of like four novels.

01:01:57   And so we built this entire pipeline that will translate using AI, our help desk, for example.

01:02:04   So now you can navigate the Obsidian documentation in 35 or 40 languages.

01:02:09   And we know that the English version is solid.

01:02:13   And we use that as a source of truth to translate using LLMs in many other languages now.

01:02:19   And we have a really great pipeline for that.

01:02:20   And then we have the community help us clean that up.

01:02:24   So, of course, the translation is not going to be as good as a human translation, but it's better to kind of have the breadth of 90% accurate translation so that it helps people, for the most part, understand what Obsidian does.

01:02:41   And then that last 10%, we can work with contributors to work on the cleanup.

01:02:45   So that's an example of like, if we had tried to do that six months ago, or let's say a year or two ago, it would have been really hard.

01:02:55   But now we can do that using these tools.

01:02:57   And anytime someone pushes a change to the documentation, we have scripts that will automatically run through the entire translation pipeline.

01:03:04   And within five minutes, the update is now in 40 languages.

01:03:08   I mean, just from the way that you've already described the company in this episode, basically, it wouldn't have existed, right?

01:03:15   You may have had maybe five.

01:03:17   Yeah.

01:03:18   Right, because you may have hired someone and they can do five.

01:03:20   But the other 30 probably would never have happened.

01:03:24   Well, you can see the before and after.

01:03:26   If you go to the Obsidian help repo on GitHub, you can see what it looked like before, which it was translated in like 12 or 15 languages by the community by hand.

01:03:37   Painstakingly, and a lot of pages were out of date, you know, things like the commercial license changed.

01:03:44   And then like in half the languages, people thought that you still had to pay for a commercial license because we didn't have that ongoing cleanup going on.

01:03:52   So I honestly think that what we're doing with the help site is probably not even done by most of the biggest companies in the world yet.

01:04:02   But it's something that is now possible and relatively easy if you know what to do that just wouldn't have been possible before.

01:04:08   So I'm optimistic about that because that's democratizing.

01:04:14   You know, this is a version of AI where it gives more people in more places more access to this tool that is free and that is empowering to them and that gives them control over their life and independence and privacy.

01:04:27   So I think that's a positive outcome, but there's a million things in between that I'm not so sure about.

01:04:32   You mentioned plugins.

01:04:34   There's a rich ecosystem of plugins that people have made for Obsidian and Themes so people can customize the app to do things the way that they want to do things.

01:04:44   But I am sure that an ecosystem like this makes the app more complicated to test.

01:04:51   What kind of work do you have to do on that?

01:04:54   Because ultimately, people are trusting you with their second brain.

01:04:59   Well, how do you work to try and make sure that the app is maintaining the stability that you want as you're making changes?

01:05:06   Yeah, this is another one where you're talking about trade-offs.

01:05:12   And I think there are two major kinds of tools.

01:05:17   There's the tools that work a specific way and don't let you modify them.

01:05:21   And you have to adapt your brain to that tool.

01:05:24   And the benefit of those tools is that they're going to be more stable.

01:05:29   They're going to be more prescriptive.

01:05:32   And maybe that helps you have a more predictable way of using it.

01:05:35   But you have to adapt to that tool.

01:05:37   Obsidian is the opposite.

01:05:39   Obsidian adapts to you.

01:05:41   You can shape it.

01:05:42   It's a piece of clay that you can modify in whatever way you want.

01:05:45   You can build plugins.

01:05:46   You can download plugins.

01:05:47   You can modify entirely the way it looks.

01:05:50   And you can 100% shoot yourself in the foot with that.

01:05:53   You can get lost.

01:05:55   All the buttons have disappeared.

01:05:57   And that happens all the time.

01:05:59   People are like, I downloaded the latest update and now I have no buttons.

01:06:03   What's going on?

01:06:04   And it's like, oh, I guess there's this one theme that was like expecting this one thing.

01:06:08   And yes, that is the trade-off.

01:06:11   And we can never fully control that because no one has to ask us permission to make a plugin or a theme.

01:06:17   Anyone can just do it.

01:06:19   And so it's somewhat impossible for us to test all the edge cases.

01:06:23   What we do is we have a rollout strategy where for every update, it goes through these different stages before it goes to a public release.

01:06:30   And usually every update spends somewhere from a few weeks to a month in this stage we call catalyst or it's like a beta basically.

01:06:41   And that allows us to hopefully find these edge cases before the update goes out to everyone.

01:06:47   But yes, it's certainly a problem, but it's a conscious choice that we're making that we want the tool to be extremely modifiable.

01:06:55   And that comes with its downsides.

01:06:57   What do you think the future looks like for Obsidian?

01:06:59   Do you have things that you're driving towards that are important to you as a team?

01:07:03   Yeah, they're still the same core things that are on our manifesto.

01:07:08   We want to keep democratizing the app.

01:07:10   We want to make it as easily accessible and private and easy to modify for everyone.

01:07:15   So those are all the axes that we're always working on.

01:07:18   As far as making it more open and democratic to people, I think a big area that we're working on and thinking about is collaboration.

01:07:25   That up until now, Obsidian has been very siloed as an app.

01:07:29   Like it's an app for you to manage your knowledge or even in the work context.

01:07:34   It's for editing your local markdown files on your work computer.

01:07:38   So this has been a big request from the community.

01:07:41   How do you, I mean, as simple as share a grocery list with your partner, for example, to, you know, more complex workflows.

01:07:48   So that's on our mind.

01:07:49   There's a lot that we can do with bases, which is the kind of database functionality on top of markdown that I described earlier.

01:07:56   So people want to be able to do some more things like Kanban and calendars and things like that.

01:08:01   So we have some things in mind for that.

01:08:04   These are all like relatively near term, like things that are on our radar.

01:08:08   When I'm thinking long term, like 10 years, I have no idea what is going to happen.

01:08:14   I don't know how we're going to be using computers.

01:08:16   I don't know if anything that we're doing is still going to be relevant.

01:08:19   I do feel pretty strongly that we will continue to have things like markdown and plain text.

01:08:26   I think that is not going to be going away anytime soon.

01:08:29   But our relationship as to how those things are created might change pretty radically.

01:08:35   We'll see.

01:08:36   We've created a team structure and working environment that allows us to be very flexible and fast when the time comes.

01:08:44   At the beginning of the year, we put out this thing called Obsidian CLI, which was a way for you to interact with Obsidian purely programmatically.

01:08:56   So through the command line, very geeky, but it turned out that a lot of people who were using Obsidian to automate things like task management were asking for this capability for a long time.

01:09:07   People who are developing plugins and themes wanted to be able to debug them or work with their AI agents to like help build the app like Vibecode, a new plugin, for example.

01:09:16   And then there were lots of people who wanted to use these like open claw type of things that are coming about and have them write to Obsidian.

01:09:26   And so since our principle is we want to stay private, we want to avoid having all these bells and whistles like invading your personal space.

01:09:34   You know, why don't we create this way that programmatically you can interact with the app from outside of the app.

01:09:41   And we noticed the demand for it really spiking all of a sudden over the Christmas break.

01:09:46   And we were able to put something together in a couple of weeks that was very functional and people seem to love that capability now that it exists.

01:09:54   And then since then, many other companies have built CLIs kind of similar to the Obsidian CLI.

01:10:00   But the fact that we have this like very flexible, no meetings, like I have chunks of time.

01:10:06   So if next week some random new thing comes about, we tend to not want to use that in a reactionary way.

01:10:13   But if we see an opportunity, we can act on it quickly if we need to.

01:10:18   So who knows what the future will bring.

01:10:20   If people want to try out Obsidian, where can they go to find it?

01:10:24   You can go to obsidian.md on your web browser of choice and download the app there or on the App Store for iOS or Android.

01:10:33   I would suggest trying it on desktop first because it will probably make a little more sense there.

01:10:39   And then mobile is sort of your companion version that you can take wherever you go.

01:10:43   And then we'll put a bunch of links in the show notes.

01:10:45   But I think try to keep it simple.

01:10:47   You don't have to go down this realm of all the plugins and all the themes right away.

01:10:51   I would encourage people to keep it simple and just start writing.

01:10:55   Maybe bring in some of your existing notes from whatever system you were using before and experiment with links and all of that.

01:11:03   And then you'll see if it's a fit for you.

01:11:06   It's not for everyone.

01:11:07   You also mentioned essays.

01:11:09   And I recommend people go read a lot of the stuff you've written.

01:11:11   And I'll put a lot of links in the show notes.

01:11:13   But otherwise, where can people find you online?

01:11:15   My website, stephango.com, is a good place.

01:11:19   And then I'm on all the different social media platforms as Kepano, K-E-P-A-N-O.

01:11:24   Thank you so much for listening to this episode of Cortex.

01:11:28   And thank you so much to Steph.

01:11:30   I loved this conversation.

01:11:31   It really is a favorite of mine so far.

01:11:34   And I hope that you felt the same.

01:11:36   If you want to get more of this, we talk even longer and more text.

01:11:39   Go to getmoretext.com and you can sign up for an ad-free, longer version of this show every single month.

01:11:47   But as always, thank you so much for listening.

01:11:49   It always means a lot to me.

01:11:51   And I'll be back next time with another wonderful guest.

01:11:54   Until then, bye-bye.