Kourosh Dini is a psychiatrist, psychoanalyst, and productivity expert. He is the author of an excellent book on how to take smart notes using DEVONthink, a personal information management tool. In this conversation, we discuss smart note-taking and how DEVONthink can help us work more effectively.

Show notes

Show notes include Amazon affiliate links. We get a small commission for purchases made through these links.

If you're enjoying the show, please rate or review us in Apple's podcast directory:

This episode's transcript was produced by an AI. If you notice any errors, please get in touch.

Read the transcript

Jorge: Kourosh, welcome to the show.

Kourosh: Thanks so much for having me, Jorge.

Jorge: Well, I'm so glad that you are able to join us. For folks who might not know you, can you please tell us about yourself?

About Kourosh

Kourosh: Sure. Most of my work is I'm a psychiatrist and psychoanalyst. And I work with clients, I see patients and do some medication, but also do a lot of talk therapy type work. I've also developed into a writer: I write about task management, I write about taking notes — basically things that involve trying to do things that feel meaningful, trying to do good work. And, throughout my life I've also been a piano player, musician; I like to tinker around with sounds. It's a lot of fun and I've yet to stop. And I can add one more: I enjoy video games. So all of that together is whatever I am. I guess that's how I introduce myself.

Jorge: Well, that's great. I reached out to you because of the productivity side of that formula. I have been using a tool called DEVONthink for a couple of years. And it wasn't until I read one of your books called Taking Smart Notes with DEVONthink, that the tool really clicked for me. And I'm hoping that we will get into productivity and more particularly note-taking. The book, like I said, it's called Taking Smart Notes with DEVONthink. What are "smart notes"?

Note-taking systems

Kourosh: What are smart notes? You know, I take the title from, Sonke Ahrens's book, which was How to Take Smart Notes. And he had based it on the Zettelkasten approach; this idea of having individual notes that really captured a single idea that would then link to other notes. Which in turn would link back and based on this approach that the sociologist — Luhmann was it? — that put together this analog system of note cards. And then Sonke Ahrens had translated that into these digital versions. So, smart notes, I think encapsulates a lot of different ideas that come from that very simple process. Again, the simple process is: You have a single note that has maybe a single idea to it, and then you connect that to other notes. And what makes it smart, I think, is where you start to reflect on those notes. How you start to develop them over time, how they start to argue with each other in time, because what you've written now is different than what you've written in the past, and you start discovering things. It's not so much the notes themselves, so much as the effect they have on you, I suppose.

Jorge: I remember when I was in school, I would take copious notes of what the teacher was saying. And I would try to transcribe things verbatim, you know, and I would always be behind the words that were coming out of the teacher's mouth. And later on, when I was in professional context — in meetings — I would also try to take notes of what was being said in a meeting, right? And I was not trying to be verbatim at that point but trying to summarize on the fly. And I'm saying that because I think that for many people, the idea of notes evokes this notion of just writing down the things that you're hearing or seeing in the environment. But what I'm hearing from you in this concept of smart notes is slightly different, no?

Kourosh: Absolutely. I mean, I came from the same sort of process of taking notes that, maybe I wouldn't write it down verbatim, but I would just try to write down whatever I could when I was in class, similar to what you're describing. But then the issue is that — at least when I was doing it — I wouldn't have a destination for it beyond maybe an exam or the thing that I was assigned to. Maybe do homework with or something like that. Because it wasn't embedded in the system that I was developing for myself — just this idea of having my own thoughts and connecting them — it didn't really prompt me to clarify my thoughts and so in that sense, the relationship that one has with their writing, or the relationship I had with my writing, changed significantly once I started to develop a system that was my own.

Jorge: When you say 'system,' I'm thinking it's not just a repository of things. It's also composed of processes and ways of making the ideas actionable somehow.

Kourosh: Absolutely. You want the ideas to be able to come to you when and where they are useful to you and you want them to stay out of the way otherwise. And to do that isn't that hard from using a system where you just... you link to things that are meaningful and to any particular note. But then as you develop that, the hard part is where you start looking at what these notes are saying and how they might be different. The perspectives that these notes have on the same object, whatever it is you're exploring, you might start thinking, "one of these has got to be wrong." Or "maybe these are both pointing at the same thing and there are different ways of looking at it, and how do I reconcile that?" Whether it's my own thoughts from the past or some other authors ideas. So, when you try to achieve the sort of coherency between your ideas, that's I think what I'm referring to when I say 'system' — that when you do that, you're trying to achieve a coherency of meaningful ideas within yourself because you're trying to understand it and build on it at the same time.

Jorge: And this coherency is something that before using tools like DEVONthink I would do inside of my head, right? Again, by writing on a sketchbook, but I was limited to what was on my mind. And the system that you're describing, at least as I've built my own, based on the things that I read in your book, is a system that augments my mind in that it takes these ideas out of my head, puts them in what is really a database, ultimately, that allows me to easily find relationships, that would not be as discoverable otherwise. Is that a fair description of it?

Kourosh: Absolutely. Yeah. Once you put it down — once you've written it in a way that's easily accessible — then the work of having to hold it in your head is relieved. So, you can actually do the other work of thinking on top of that. You can build on top of those ideas much more easily.

Why DEVONthink?

Jorge: So there are several systems... several tools let's say so that we don't confuse folks by over using the word 'system.' There are several tools that can be used to implement such a thing. I was in a discussion a couple of weeks ago with friends who were talking about migrating from Evernote and they were considering Notion. Or another one that we hear about a lot these days is Roam Research. And I'm wondering, why DEVONthink?

Kourosh: You're right. There are quite a number of note-taking apps and new ones coming up all the time. DEVONthink... so I've been using it for several years already. Now it's been probably at least a decade that I've been using it. When I first approached it, I was kind of using it as a Finder equivalent, just throwing things in there. And there were little bits that had some benefits to it. Like, I could link to anything in it and-it was a strong, good, robust link. It wouldn't break down like some of the Finder ones and the alias function, which in DEVONthink is called 'replicant' also was more reliable. It was good. But I didn't use it too much beyond that. And then once I started to do notes, certain functions in DEVONthink became much more apparent and powerful. So probably the biggest example is the AI. One of the things that distinguished DEVONthink I think head and shoulders above just about any other a note-taking app is this AI. And at first, I thought of it more as a gimmick. I didn't think of it as very useful. You know, you throw a bunch of PDFs in there and maybe one of them it would say, "Hey, what about these other PDFs? Are they useful to you?" And, I said, "Okay, yeah, that's nice." But when I started to take these notes, and when I started to organize it myself, that's where the AI started to, I guess, rest on my own organizing process. So, now when I write something down, let's say in some particular niche of psychoanalytic thought, or maybe I'm writing about, you know, I've been interested in; structure of stories, I write some small nuance of that. Suddenly in the sidebar it shows me a handful of ideas that I've already written that could be related. And it's not that it's just taking the same words or something. It's not just saying, "Oh, I've mentioned the title of this somewhere else." It seems to go through this process of thinking about the relationships of the words together in such a way that it feels meaningful. It feels like... like if I start writing about character, then I discover ideas from stories and how characters are built on story, but I can also have it present things about defense mechanisms that might be more relevant than psychoanalysis. And suddenly I can think about these two very different approaches to the idea of character and see where they overlap, how they go together. And, you know, oftentimes I might think of these sorts of associations myself, but it's very nice to have the system say, "Hey, these are other things you've written that may not seem directly relevant — you may not think of them immediately — but hey, you might think that they're relevant." And very often they are. And it's just so lovely to have that. So, that's one — I'd say impressive to me — reason to stick with it, but there's others. I mean, I can throw any file in there. I can have audio files. I can have image files. And there are tools that work with these within DEVONthink, as well as the files are directly accessible by anything else. It's not just, I can export them. You know, I can do that. But I can also open a particular file at any point with any app that would work with it. So, a text file... I have half a dozen note editors that whichever one I feel like working with because one's better than the other at something... I can do that. Save it and jump to another one very quickly. And they're all sitting happily in DEVONthink where I may have tagged it, I may have linked it to who knows what else as well as multiple databases. So, anyway, I can talk about that too. Point is, there's a good number of reasons why, to me it just reigns supreme in terms of these note-taking apps. I will say that there are some of these other apps do things that DEVONthink doesn't. Such as, you mentioned, Roam. Another one that's come out recently is Craft, where you have these, blocks, these block references. And DEVONthink does not do that. I've tried them out, and I continually stumble on myself, trying to make them work. So maybe that's part of my issue. But in the end, I've found that I very much value a simple text file. There's something about it that feels more paper-like, that feels more direct. And I don't mind rewriting if I need to, though I don't actually find myself doing that very often. So, in the end, DEVONthink really is the powerful tool for me.

Jorge: I haven't played with Craft, but I did play a bit around with Roam. And when I hear you talk about blocks, I think that what you're referring to — and I just want to be clear on it for myself — is the ability to treat elements of a note or a document that are more granular than the note or the document itself... treat them as individual entities that you can point to and manipulate somehow, right?

Kourosh: Exactly. Yeah. That each line can be changed, adjusted, can be referred to — some of them in quite powerful ways. And you can have combines and you can have images placed there and you can drag and drop them around. And yes, refer to one particular line in a particular note, from any other note.

Jorge: And the trade-off there to your point when you're talking about the paper-like experience and also DEVONthink's ability to host files that are openable in other applications. I think that one of the trade-offs there is portability, right? In that if you have a system that lets you deal with elements more granular than the document, all of a sudden you develop a dependency on that system.

Kourosh: Absolutely. Yeah, no, once you do that, you're somewhat fenced in. Even if you can export it. Even just psychologically, you get connected to that system. I would rather have a tool that lets me manage the things I work on external to that tool. You know, if I have a bunch of nails, I don't want to have a certain brand of hammer that only works with those nails.

Jorge: Right. And to illustrate for folks listening in one of the things that I learned from reading your book, was how to deal with the notes that I'm taking in DEVONthink as markdown files, right? Markdown being this markup language that works on plain text files. And I can use BBEdit, which is my text editor of choice, when working with DEVONthink think files and there's this portability that happens not just... not just portability of the entire set of notes, but even when working day to day with the thing. It encourages you to somehow use other elements that you're more comfortable with, or that may do a better job than DEVONthink itself for whatever task you're trying to do, right?

Kourosh: Yeah, absolutely. I think that's exactly it. Like if you like BBEdit, it's fantastic. You know, I like iA Writer, is one. I like type Typora is another. I can jump between a number of them and just, they all work.

Kourosh's workflow

Jorge: I'm wondering about your workflow when taking notes. When you were describing it, you were talking about discovering DEVONthink's AI, surfacing links to notes that you had taken previously. You also talked about PDFs. Are those PDFs of things that you yourself have written or PDFs from third parties, journals, stuff like that?

Kourosh: Both. Whenever I find a journal article, for example, that I want to add to the system, I'll add it to a folder titled 'Reference.' And I might even put that into sub folders that it relates to. Or anytime I complete some major project that I'm writing, even if it's based on stuff that I've done within DEVONthink — the notes that I've written there — then that complete article that I've written is now a reference that I can use. And I'll add that to DEVONthink. I think actually now that you mention it, I think that's the other part of DEVONthink that I didn't mention that I really liked a lot is moving from notes to completion, to a complete something. I know it's a little tangential to what you just asked, but I was able to take, you know, about 30,000 words of notes — over 300 notes — about... you know, as I was writing about, ADHD and the psychodynamics of it. And I read a bunch of papers. I imported them, about six papers to start. Followed their references, went to about a dozen, had maybe 20-30 sitting there that maybe I didn't read deeply, but at least a dozen that I did. And I was able to take those 30,000 words, 300 notes, drag them into Scrivener. In Scrivener I could, play with the corkboard there and arrange them nicely, you know, in the sort of bottom-up organization, where I discover, "Oh, this kind of goes here, this kind of goes there," and, figure out a good flow of where the words would go. And how I could... how can I lead the audience? And, in the end, I edited it down to about 18,000 words, which turned into a four-session lecture. And it wasn't hard. It was enjoyable to go through that process. You know, to discover along the way as I created this final piece. Which then I took as a PDF, and actually as a Scrivener document, and put it into my references so I could connect to it and link to it again, further, in DEVONthink.

Jorge: The way that I'm hearing that workflow works is that DEVONthink is the system where the knowledge is stored in a way that allows you to easily surface connections with other pieces of knowledge that might have fallen off the table or been something that you collected a while back. But then the actual process of creating a new work based on those connections happens in another tool. Is that right?

Kourosh: If I'm creating like a final piece of something? Yes. Like if I'm aiming for whatever the medium is, it's going to be outside of DEVONthink. So, if I'm thinking of a Keynote presentation, I'll use that. If I'm thinking of a long form text, probably I'll be using Scrivener, but absolutely the consumption, digestion, working-through of knowledge and the accessibility of my ideas, all happen in DEVONthink.

Jorge: Yeah, I'm asking because that's something that I'm struggling with myself. I'm always facing the question, should I keep writing this note in DEVONthink or do I need to move it to Ulysses? Which is the... it's what I use instead of Scrivener, it's the more, kind of long form thing. Or should I do this one in BBEdit? And it, it feels like part of the deal that comes with a powerful complex tool like DEVONthink is that by opening up so much choice, it does become a little complex in that you have to make choices about what you're going to do and where.

Kourosh: I would divide it as... like, I have a sense or a feeling of what I want my DEVONthink database then notes to do... like it's a search of knowledge. It's a development of knowledge. It's a growth. And, if I feel like the words have a destination, let's say a post or something like that, I might... I like the idea of a singular idea as being a note, you know? Trying to get each note to have a single idea. And as long as I have the single ideas represented in my database, DEVONthink, then I can take any of them and weave them together into something longer form elsewhere. So, if I start writing something and I'm wondering, "should I start writing this elsewhere?" The only thing I have in mind is, "well, are the ideas represented in my database?" And if they are already well then, that's great, then I don't need to edit for some flow between the ideas necessarily, that might be more aligned with whatever its destination is. And that's when I might take it out. And if I discover new things as I write that, then, you know, I'll throw them into the inbox and DEVONthink can work on them later.


Jorge: One issue that I wanted to discuss with you, and it's just because it's something that I'm using right now, an aspect of DEVONthink that I'm using right now, and I wanted to touch on it because I'm finding it incredibly powerful and feel like it's something that folks would appreciate hearing about. Like you're saying, I'm working on something right now where I have a final destination in mind, in this case, it's a set of Keynote presentations. And what I'm using DEVONthink for is making these connections between ideas and discovering connections that I might not have been aware of before. And I, like you were describing, I've collected a lot of my own notes, a lot of PDFs, bookmarks to websites and I've been tagging those things as I import them or create them in DEVONthink. And then I have set up smart... I don't remember the right terminology, but it's like the equivalent of 'smart agents' in DEVONthink that surface the items in the database that have that particular tag. And what that's allowed me to do is to very quickly discover these relationships that I have been slowly accumulating over time and — there's a question here, I promise! — The question here has to do with tagging as an activity that you do at the moment of capture versus tagging as something that you do at the moment of reflection. Because my ability to surface those items is going to be dependent or greatly improved by having good tags. But sometimes when I'm in a hurry, in the moment, I might tag something with one or two tags, but that might not be rich enough to describe the full utility of this idea, right? And I'm wondering if you have suggestions or thoughts about this relationship between bottom-up tagging in the moment versus the more reflective structure that happens when you circle back to add meaning to things.

Kourosh: Yeah. So, most of the way you described it, I think it's similar to way I might do it, which was: If I have a particular project or something that I'm working on, and there are notes, ideas, that are related to it, I might give it that particular tag. The second way you described it is I might tag something with multiple tags and those multiple tags may not fully describe everything about it. That second way I avoid. Any tag that I have, I've made it a principle for myself to have a very clear purpose. I think it's often approached... and I don't know if I'm misinterpreting, please let me know. But I think it's often that tags can be approached as like, "Well, I'm going to add everything that comes to mind about it." Like, it's used associationally, and then hopefully you'll be able to discover that later on in some association with whatever. But I've very rarely found that to be helpful to me. So, instead — and that's maybe partially because I've come to rely on the AI in DEVONthink — that I would much rather just have... Let's say I'm working on the ADHD idea. I have a tag just for that — in the psychodynamics of ADHD, that was one tag. And everything that related to it, got that tag. And then later on, I realized there were certain ones that I thought would be important to have and I'd forgotten to tag them. So, I created a smart rule that said, "search for everything that has the phrase either 'ADHD' or maybe the phrase 'concentration', or whatever it was, and also does not have that tag." And I was able to search through and then, "Okay, these are the ones that need to be tagged." Okay. So, then I go ahead and tag them. And then once I have them all tagged, now I have all those notes. And that's where I can grab them all, drag them into Scrivener and do whatever I want with them. Anyway, the one question you'd said was, "Do I tag it before or after, as it enters, or later on?" I'm not sure it matters. I think, whatever... when you realize that it's a part of your project, that's a good time. You know, I work to have it so that everything's within the notes and not in, PDFs or scraps or webpages. Once I've fleshed out all my thoughts and now, they're all notes that are interlinked, that's a great time to move it on. But yeah, I would avoid the kind of associational tagging. At least that's the way I've done it.

Being deliberate

Jorge: If I might reflect that back to you, and just as a way of starting to wind down the conversation, it feels to me just from hearing you describe it, and from my own experience, that systems like DEVONthink are most useful when they're used purposefully, where it's not like an arbitrary dump. We used to have this term: a junk drawer app, right? Like, where you just dump stuff. And it doesn't feel like that's what this is. This is really a purposeful thinking tool. And if you bring purpose to it, you're going to get a lot out of it.

Kourosh: Absolutely. I will embarrassingly say, though, I do have a database in DEVONthink that functions as a junk drawer. So, I'm not immune to it. But the database of my notes? That is very deliberate. There's another database, which is a bunch of websites of "I found something funny," or "there was a nice joke," or "there's some social-something happening." And that just... I have an organization in there, but I have yet to figure out what I'll do with that organization. So, it's a junk drawer. But I don't get much out of it unless I'm doing it like I do my notes. The notes? That's where it becomes powerful.

Jorge: My dream is for the junk drawer aspect of this system to serve up serendipity somehow.

Kourosh: Sure! You could make that happen, now that I think of it. What you could do is you can have your notes database open, and then you have also the junk drawer database open, and then as you're working, consider also — see also — all that... brings anything to mind from in DEVONthink. It'd be an interesting experiment.

Jorge: Well, I'm going to try that out. I frankly didn't even know that that was a thing. I thought that databases were separate.

Kourosh: Yeah, you can do it. I'm pretty sure you can. Now that your question and I'm like now 95% instead of a hundred percent certain! I have to go double-check now. But I'm pretty sure you can do that.


Jorge: Well, fantastic. This has been such a pleasure talking with you about this, and I feel like we could keep geeking out on this. Where can folks follow up with you?

Kourosh: Sure. I have a couple of sites. One is beingproductive.org and that's where you'd find the things that I write about in terms of productivity, in terms of note-taking. I write about the use of the task manager OmniFocus and I also write about just being productive in general, without any tools. What does that mean? And then if you're interested in more of my you know, other interests like music and games and psychiatric type things, that's at my... just my name, which is: kouroshdini.com. Which is a kouroshdini.com. And that links to basically everything that I do.

Jorge: Well, great. I'm going to include links to all of those in the show notes. Thank you so much for being with us today.

Kourosh: Thanks so much for having me. I really enjoyed our talk here.