Select text to annotate, Click play in YouTube to begin
00:00:02
What a great tour the summary can be found in the timestamps for the video, but I'd like to point out a few things. Through Maggie's daily tag. She has a field for what's on her mind. This makes it really easy to scan through the list and see how that's changed. Over previous weeks, we show an advanced way to set it up, which is how she originally set it up as well as a simpler way to capture similar benefits.
00:00:27
Her writing pipeline is gorgeous. You know, generally what she's doing there. Tagging, any writing idea she has with a pattern or garden note. These notes can come from anywhere. Maybe the idea came up in a meeting maybe while taking notes on a source or maybe while journaling. These notes have a field for stage think to do in progress published and motivation level. So she writes up a query for nodes tagged with garden note or
00:00:55
pattern displays as cards, groups by stage and sorts by motivation. So you'll just have to see it to get a sense for it, but you could use the same workflow for writing pipeline, for task list, for job applications, whatever. She also likes to use fields as prompts. She does this to think through decisions more clearly, as well as to write better introductions for blogs and presentations. On top of her workflows, she likes to use tags and fields for just about anything.
00:01:24
For example, travel plans. She's excited about the addition of ontologies into tools for thought, allowing Tana to merge a graph based outliner with structured objects that have types as well as AI augmentation, similar to her work with illicit. She suggests that new users start with a project and think about what they want to get out of their system without any further ado.
00:01:49
Here's the first of many Toni tours. For a little bit of background for people. Like, who are you, what do you do? Why do you use tools for thought and personal knowledge manage? Yeah. So hopefully this is on the video, but my name's Maggie. I, I am about to start working at a company called or which makes a product called Elit, which is machine learning and natural language processing to help academics and researchers essentially be more effective and
00:02:18
more efficient in their research, help people do open-ended reasoning. Mm-hmm so it's a, it's a research lab very much like kind of just exploring that space. Yeah, I've been. Involved in whatever we're calling it tools for thought note, taking personal knowledge management, like every buzzword for like two and a half years, like I was very early to Rome research as we, you, I feel like we started the exact same week was what we found out. Right. And I used Rome for, for two years, like intensively.
00:02:48
I mean, it was one of those where like, when I would track it, right. It's like at least three or four hours a day. I just, it was just my primary database used it just. Mostly I, I do a lot of writing online. So a lot of my knowledge management systems are focused on a very specific kind of output, which are essays and kind of design patterns. I, I write about a lot, so I I'm a UX product interface designer, like some mix of those things.
00:03:13
Mm-hmm so I write about that field in specific visual programming developer education, like interface design. On my personal site. And so all the things I do in note taking apps are directed at that as well as professional stuff. So like research that relates to my job, trying to problem solve issues at work. Those are kind of my two main use cases for it. So my system that's like designed to serve those needs and not anything else.
00:03:41
I'll say mm-hmm mm-hmm yeah. Yeah. That makes sense. And what would you say attracted you to, to like, why is it like you used Rome so much? Why is it that like Tana is now, you know, scratching enough of an itch for you that you've been like using it as your main driver? Yeah. So even from like the very beginning of using Rome, there was a core piece missing.
00:04:10
That I'm kind of surprised it has taken this long for like note taking tools to catch up with and for people to implement mm-hmm and that is ontologies essentially. Or we could call them schemers, but essentially structured types in a system. So a book type, a person type a company type. There are a couple of apps doing this now. And to was when I first saw it, when I realized that it was so structured around having these ontologies or types that you could.
00:04:35
That was what made me be like, oh, we wanted this the whole time in Rome. Like we all tried to do it with attributes and I had this like really hacky way of trying to put the type in the name of the, each like page node. And you could never really query it nicely the way you wanted to. Yeah, you could never really like properly use it. It was like trying to implement an ontology in an on system that wasn't designed for ontologies. So from the beginning, that was like a frustration. And my, my last job was with hash, which was a company, which is working
00:05:04
on an ontology based block editor, like a knowledge management tool. And so that really got me invested in it as well. That made me start thinking much more about ontologies and the semantic web, like that legacy and how that might influence note taking apps. So that became like, I was just then became obsessed with looking for other companies that were trying to do the same thing. And then yeah, when I found town, I went, oh, this is a really well implemented version of this. Cool, cool. Well maybe then let's just get into it, you know, and and take a look a little
00:05:34
bit at what sorts of ontologies and types you've incorporated into your own system. Sure. Use that. I'm gonna share a screen here. All right. Are we seeing Tara mm-hmm okay. So for people maybe who don't have context, I shall go up to my main graph like homepage. It's, it's a bit different to, to Rome, right? Where Rome is like, everything is the daily page and there's nothing else, except for, unless you really did work in like individual pages.
00:06:08
I like that Tana has this like homepage, which is kind of your main entry point, but then you also. Pages for every single day that you, so I'm gonna go to today and actually have to do the switch calendar thing, cuz I'm doing something weird with, with graphs here where this is technically my main graph, but there's a thing it doesn't do that I want it to do, which is why I'm in the secondary graph down here. But this is my main graph. I can explain that later. And the way I have my, so this is like a day type that you can can take. They're called super tags and tar.
00:06:38
I'm just gonna move the like zoom Chrome out the way. Hold on one second. And you can say what you want, your, sorry, I've lost my mouse. This is going well, there we go. You can say like, these are default content that will appear on every single one of my day pages. So I use three basic ones. I have a time log where I just note down and incredibly loose ways. What I might be doing that day. If anything, mm-hmm, , it's, it's a pretty, I'm on holiday right now. So there's very little in here, which is great. A very quick what's on your mind, which I have, I used in Rome.
00:07:10
This is essentially what I did in Rome too. So I just forward this over a very high level. Like, what are you thinking about, which I find that useful to be able to. Look at later. So I have this page here called what's on your mind. And what this does is this is a view of all my days so far and what I've written and what's on your mind. So then I can kind of scan it and be like, what have I been thinking about? Which is like, always very useful. Oh, that's super cool. Can you just like real quick, just tap on that live query right there.
00:07:39
Just so we can see what's behind this. Yeah. O okay. Yeah. So this seems like. Fairly simple one. So has tag day. So, or yeah, what what's happening here? Yeah, actually, I had never clicked on this before, so I'm not sure why. Oh, I know what I did. Okay. So this is saying everything in and has to be true, right? So I'm actually not sure what has tag is.
00:08:07
I might even get rid of that one. I'm not, I'm trying to remember when I wrote this could have been eight years ago. It is saying that I have actually defined what's on your mind, so, okay. This actually is gonna get us into like the, that let's go back to Dave for a second. That this configures see the tag. Sorry, I'm getting into the deep stuff quick. I'm sorry. Hopefully people have watched tutorials. Sorry, what? Oh, that's fine. I can set. Okay, so actually let's, let's slow down there. I just hit command. K, which opened up this? What, what do we call this?
00:08:38
The command mind menu. Yeah. Yeah. Yeah. So one of my favorite, favorite, favorite features int is I I can, right. So I can define default content on this page, but I don't always have to fill these in, cause I don't always want to fill them in there. Like the vast majority of the time. These fields that I put on any tag are entirely optional and I like to be able to conditionally add them. So I have this thing where I can set the hide field conditions to say, never show this field.
00:09:04
Always show it. Only show it when empty or only show it when not empty. So I use hide when empty for almost everything. Yeah. And then when I want to add it, so actually if I go to like, let's go to, if we go to today and then we go to tomorrow. Right. So here, this is what it looks like when you haven't added them yet. Mm-hmm so tomorrow I'll click that. And like, so that brings us back to the what's on your mind thing with the live query, where only if I have added what's on your mind with that little plus.
00:09:34
So it is defined on that day. Will it show it otherwise would have a bunch of blanks? Yeah, yeah. Yeah. What, what what's interesting is I, I think that you could. Just have the what's on your mind field and defined there in that query. And I think it would pull up the same results. Let's try. Okay. Okay. Cause actually that other stuff might be junk. I'm trying to remember what in the world I was doing. I was probably trying to like hack around with things that I
00:10:02
shouldn't have been attempting. What's on your mind. So let's say let's try one once and see if this is gonna do it done. Actually, I'm not seeing that. Oh, I'm seeing some blanks. Oh, those seem to be a bit maybe those are errors. there's, there's some funny stuff like in here, like, I'm not sure this, but sometimes you end up seeing like bits of the underlying system rise up, which is totally fine. Like we actually seen the day tag shop in here, but like that one's empty.
00:10:32
So like that might have been one Yeah. Actually, no, I think it is great, cuz like I'm now seeing this big gap in between these two days and I'm betting it's because the days before this did not have what's on your mind defined mm-hmm now it's almost every single day. Okay. Maybe I've been writing it more than I thought don't always fill it in like this one's empty. Yeah. Yeah. So even with that simple query, you know, even with that simple query, you
00:10:59
can get something like this, you know it doesn't need to be complicated, right? Nope. Although, although maybe, maybe you were playing around with something at the time. That was, that was interesting. We're actually going to this'll be coming out. Fairly soon, but we're gonna be adding like just simple filters on the table views. So like, it should be pretty easy for you to just say like, like to filter out empty.
00:11:26
What's on your mind field notes. Cool. Yeah, I think I have it more often with time log. So this is similar time log is one of the things I have on my day page. And there they'll be empties. I bet you, that's not the only empty though. It seems to be filtering out ones that aren't empties. So, so that's your day, you and, and you can like, what's cool about that is, yeah. You can like very easily pull up these, like, create these like tables out of, you know, like what, what was your time log just across, you
00:11:55
know, your entire time using Tona? Just by having a consistent ontology, I guess, guess for, for your day? Yeah. Of fields. Yeah, I'll say the, I'll go to show this one. That's like why I love these dynamic views. So no context. So this is me tracking my writing which is again, like. I, I would say at the moment, the focus of what I'm doing int only cuz I'm on between jobs for a month before I start my new role.
00:12:24
So I was trying to like get more writing out in the meantime of all these things I have in my backlog. So I write two types of things on my website, garden notes, which are just kind of like essays or notes. And then what I call a patent, which is defining a very specific design patent about something. Yeah. So like I just published one yesterday called folk interfaces, which is Sorry. I'm probably going too fast. Yeah. no, this's fine. Keep going. Okay, cool. So yeah, a patent I also like to, you can put like definitions on things int, so
00:12:56
I have this thing called patent and it says a patent is an organically emergent solution to a common design problem. This is like a Christopher Alexander thing. And I write patents that I find on my website. And so I have these like kind of elaborate list of fields on this. So I have a creation status. So like, is it an idea? Am I ambiently researching it? Am I actively synthesizing it? Or is it alive in public? Which means it's on my website and I like will continue to tend it over time. Mm-hmm so this one's an idea.
00:13:23
Stash. I have a public growth stage. That's like once I have put it on my website, I then have a three level ranking system saying like, if it's a seedling, it's like, I've just kind of like chucked some stuff up. And it's not that well done. Butting is like half grown at evergreen. Right? I'm like, this is pretty good. Like this. Like pretty much done. I have a per personal motivation thing. So number one is like, I was born to write this. Like I have to write it. Number two is like, that's a good idea.
00:13:50
And three is like me. I like not really sure about this one. mm-hmm and then I, in, in these ones, I don't have any like default data filled in, but I, for every patent, I wanna be able to make sure I. It's probably easier to look over here. The problem. So what problems consistently emerge in this context that like leads to this pattern? What's the context it is developed in related to as a, as a field they have on
00:14:19
almost every single type I have, which just says, what does this relate to? Yeah. That could be concepts, evergreens, other pieces of writing people, companies, whatever you want. Mm-hmm Yeah and solution. So I put that at the top. So like this is the solution to this problem and in this context of a pattern and then I little bit of tracking things like research notes, main points, and like things to research in here.
00:14:47
Mm-hmm also, so can you still hear when my audio is telling me that it's being strange? Yeah. Your, your audio sounds fine. OK. Cool. Sorry. My pods are like infamously crap as everyone's. Yeah. So one of the things that I'm already seeing that just like is cool about this, right? Like you can create tables or. Con bond boards out of like any of any of this data, right?
00:15:20
Like, so like you can, so like, if you want to figure out like, okay, what things should I, should I be writing about, you know, you can just like create a little query for just personal motivation, you know, and then have it grouped by the result. And then. Right there. You just like, see all the things that you are best suited to write. like, yeah, I do that on the, sorry.
00:15:47
We got here from this page, which is me tracking my process. So I have it grouped by stage of like, which ones. So I should be paying attention to this. This is like my actively synthesizing stage. I also have be sorted by personal motivation Fest. So I just reapply that sort. So you'll see here, like, oh, I was born for this comes up first and then we get into good ideas and their ideas. So this kind of goes down the list. So like things that are way down here, I like might never write
00:16:13
there, but like things up here I should be paying attention to. Yeah. Yeah. So I, but then we can do the grouping by personal motivation rather than creation status. Mm-hmm so that we get those. So, this is where the, I really love these. Like, you can show this whole thing as a list against a group. You can show this whole thing as a table which like exposes a lot more of the information. It'd have to look hide a bunch of these things for this actually be useful. But I like the card view personally, it just like helps with grouping, but
00:16:45
this dynamic grouping's really nice. Right. I can do it like public growth stage. Right. So I can kind of see, and I filter out evergreens on this too in the, in the live query. Cuz if it's evergreen, I have so many of them, it just kind of clutters up board. I'm not, I'm unlikely to touch them. So I just felt that those out mm-hmm that those would be that, but this like gives you a nice dynamic overview of kind of like the status, civil, all your pieces of writing. mm-hmm yeah, that's really cool. And, and like, I think that this just, it shows one of the key
00:17:19
strengths oft of just like, yeah. This combination of like the graph based structure alongside structured data, because like, once you add this structured data to, you know, your nodes Then that allows you to create these sorts of like dynamic visualizations for you create like a contextual dashboard that lets you like, not just have the information, but
00:17:46
put the information through a workflow. So, so I'm curious, like, okay, if you expand out one of if you expand out one of these notes what do, what do you. Sure. So one thing we can do is we can open it in line which is kind of handy. Like now we are essentially like on this page, which I could click on this and we're like on the page and can open it up and like see all the content.
00:18:15
But I can also just kind of open it from here. And see all the same stuff even in line, which is a bit crazy, which is really cool. You don't get to see backlink references, which I do use a fair amount. Actually just minimize that sidebar. Oh not sure what happened now. It's okay. So one thing you can do is if you go to settings we've got tunnel labs. We have reference count.
00:18:44
If you show, show, reference, counter all pages then that will add a number next to every node that has reference more than one reference. And you can click on that and it'll show you the back references. Yeah. So I, it's funny, I have that all the tunnel lab stuff enabled on this instance in arc because we opened it grave. Yeah, it undid them. But I do use that a lot. Uhhuh. Uhhuh. Yeah. We're just gonna get it to, to reload now.
00:19:13
Yeah. And maybe go into, you wanna go into full screen maybe? Oh yeah. Let's go into, yeah, let's go into this one. So this is a least I haven't written yet, but that I've like made a whole bunch of notes on. And I've made like a list of like research I need to do. Mm-hmm and this is where the back links come in. This is like what Ropa ended and what everyone ended up copying essentially Um-huh, but it's, it's really valuable. And I really like that, like the way Tara does it, because you have
00:19:41
fields and tight, the back links are not just a giant dump list of back links, which was always the problem. As in Rome, as you would end up with like 150 back links and then. You kind of were like, well, I'm not quite sure what to do with all these. I certainly can't read them all. You'd have to like pull out information up into the main body. It was kinda tricky to work with even just like the color coordination of these and like section them out helps me a lot.
00:20:05
So I have this field called source force. So whenever I have like an article mm-hmm, I have a source for, and I it'll be like a source for like maybe a garden note and like a pattern and I can put those in here. And then when I go to those page, At the bottom, I can see everything that is a source for this piece, which is really handy, which is different to related to which could be like tweets and other concepts and, and
00:20:29
other things that are not necessarily articles and academic papers. Mm-hmm , mm-hmm so as a result, it ends up being like little bit higher signal a little bit higher signal, too. And, and, and what I really like too, if you scroll up a little bit What I really like too. And just what I've noticed in my own experience using this is just having these fields and using that as a big way that I relate concepts together has
00:21:03
mm-hmm, created something where just like navigationally, it's easier for me to get from where I am to where I want to. Yeah, well, honestly, I even love it just as a prompting device, even mm-hmm like if I go to my patterns page and like, let's say we, we wanna define a new, like design pattern here. Right? I don't, I don't know what we're gonna call it. Colored tags. It's not new, but whatever that's on new design pattern, it immediately prompts
00:21:31
me to like, decide what my personal motivation level is for this mm-hmm Define what the problem is, define what the context is. Define what I think the solution that this like describes is like, it makes me, or like real world examples. It makes me fill in these fields. Mm-hmm so it's like giving me a template to think within, which is what I find the most useful about any ontology based system is it just gives you that structure upfront. So you're not always doing the labor of like, okay, what do I need to
00:21:57
think about again, if I'm doing. Yeah. Yeah, totally. And, and in that way it can help you think through things more clearly too. Yeah. Yeah. Like I did an experiment with introductions, like mm-hmm so I made this field for introductions and I was like, well, how would I, how could I help myself get better at writing introduc? And this is, and I had read a paper, which I really loved, which I think I might have.
00:22:27
I wonder if it's still in my room or if I did port it over yeah. Problems, rhetoric So actually, so actually we got, you are seeing here too, is also like these are back links to my, this is my row graph that I have imported in a separate graph over here that it's able to reference. Sorry, we're getting deep. So this is showing me notes from this other paper. That's about how you define problems to make them interesting with this paper.
00:22:54
And I use the structure that I learned in that paper to talk about writing introductions, where like every interesting problem starts with a destabilizing condition. Mm-hmm That destabilizing condition happens within a context. It has consequences. It has an assumed audience. Anyway, it's like this thing where I thought about how I might set myself up to write a really good introduction is by defining these fields first. Mm. And by making a tag int that enables me to then be like, oh, okay.
00:23:19
Like, how would I think about the context of this and what the consequences are? Mm-hmm mm-hmm. Totally. Do you find room for freeform fi freeform writing within that? Yeah. So this was very much just like a more structured approach to it that I would say I wouldn't, it's not like I like whip this tag out every time I wanna write an introduction. I mostly don't. But if there's times where I haven't written a good introduction
00:23:46
and I like don't know what to. Pulling this out feels like pulling out some sort of like prompt card or something, or some sort of useful tool that I can call when I need it. But it's definitely not like I use this every single time. Yeah. Yeah. One, one of the things that I'm just noticing about how you are using Tona, that I just think is very cool. And I think other people should note is that you're really like.
00:24:14
You sort of like, started with thinking, like what types of things do I have in my notes? You know? And, and, and then, and then from there just like creating those tags, a lot of this structure like ended up like easy to create, you know, like, like if you click on one of these tags, like it already just generates a query for times where you've used that tag, right?
00:24:38
Like, like even, just, even just going from there and, you know, There is a little bit of a difference. You know, like with people like you or I who have been using tools like these for years. Right? So like, so like we, we have. So like going into it, we probably already have an idea of some of the things that like, we want to turn into things, right. Like for me, I knew going in that claims questions and decisions
00:25:08
and sources and people, those are all like things that I use that have relations between each other. Right. Yeah. So I went in and just created those tags. And then from there it was like, the structure was fairly emergent, you know, and. But, but like, I'm curious, like what advice you would give what advice you would give new users to develop a system that works for them. Well, this is interesting is there's a page here called draft on that I
00:25:37
made on the, like the first day that I got into tar and like poked around and realized what it was, and this hasn't ended up being like the exact pology I use, but it was really helpful for me to sit and write out this list of where I, like, I was like, okay, I finally have an oncology system. What, what things do I make? I know I want, yeah, right. I have people profession, software organization, mm-hmm design principle, you know, these things community. Most of these things have held up on my scheme as page you can look at after it
00:26:05
like shows what I actually have, but on each of these, I kind of started drafting out what my relationships were gonna be. At least didn't exactly translate over, but it was really helpful for me to attempt to do this first. And then. My actual scheme is page. So this is scheme is like lists out all the actual types that you have in your graph ended up being much more organic and emergent, which was kind of the plan. It was like, I got started with a few that I thought, and then as I
00:26:34
went, there were moments where I was like, oh, I like need an airline type because I'm like tracking flights. And I like need to know every time I like fly with Virgin. So I like see all my flights down here. Like this is really useful. Like when I'm, I've been flying a lot, this. But even just being like, did I, did I like buy a check bag? Like I forget, like, what was the, yeah. Like, you know, like what was the flight number again? Cuz I need that, like this became really useful. right. So there's things you don't expect that you're gonna make types for and then
00:27:06
you end up being like, oh yeah, like I, I have a bunch of strange ones in here house for rent house for sale. Those end up being relevant at some point, you know, mm-hmm conferences, podcast, episodes. It's really fun that I also like how easy it is to make a tag and there's no cost to it. Like I don't, yeah. I see it as like it cluttering up my system. Like there's just things where I'm, I'm like, well, I'm,
00:27:31
I'll only be able to use this. Like once to like, like, like how many times do I really am like buy a house. I will only use that tag, like for one project, like the attempt to buy a house project. But it's useful for that. And it can just like sit there and it doesn't like invade the rest of my. Right. Right. That's really cool. I'm curious about like some of these other you know, like little use cases that have popped up, you know, like, like just flight tracking.
00:27:59
Like I can immediately see that being useful and that's not something people usually think of when they think like, oh, tools for thought, settle, casting, evergreen notes, like whatever. Right. But like that coexist, you know, with, with your system. Right. So like, I'm curious what sorts of other things COEX. Interesting. Well, I haven't okay. I haven't used decision as much as I want to, but I know you on this. So I would love to kinda like hear or see how you're using it.
00:28:25
At some point, I don't have many in here, but it was really useful for really stupid small ones. I had a debate with my partner about whether I should buy an ice cream maker, but it was a good, it was a good test of this decision tag. So I went, okay. What's my reasoning. Have some research notes on. I have like a working hypothesis. So like my best current answer to this, to this decision question, right? What are the costs? What are the benefits? What are the known unknowns, which are in themselves questions, which
00:28:52
have their own working hypotheses in them and like reasoning. So like even this like silly kind of example, I found it really helpful to just like think through it in this structured way. Yeah. Some of them are bigger. Like, should we buy a house? Like that is actually really useful. And I like had to do a bunch of thinking in here about like how much it costs to like you know, mortgage rates and like, mm-hmm, what the average prices are around us and how much you pay just brokers to like,
00:29:19
get the whole house, like even set up. Anyway, the current working hypothesis is, is not right now. But this was super helpful to think through all these tiny, like questions on like yeah. There's so many smaller ones in here. Like what taxes would you pay? Like would it be cheaper to buy and rent over a 10 year period? Mm-hmm anyway, that was the one that like, I didn't go into tar being like, this is gonna help me answer. It's kind of like outstanding logistical questions about my life.
00:29:47
Right. But that turned out to be really great. yeah. Yeah, totally. And like, like this here, like it's not something that. You know, like viewing a table of all your decisions and seeing, should we buy a house in London right next to, you know, like some product decision for something you're working on for Elit like, like that, that isn't necessarily useful, but like, what is useful is thinking through this in like a structured
00:30:17
way, like just the way that you. You have this hard decision where there's like a lot of things that are like, kind of fuzzy and just like figuring out, okay, what are the com what are the component questions that I need to answer in order to get to the point where I feel like I have a good decision here, you know? And and breaking it down that way is valuable. Yeah. Like I, yeah, like I mentioned, I haven't used decision as much, but questions.
00:30:42
I find ed questions in my old pro graph that I used extensively. So those poured over quite naturally. Yeah, but it's, it's the same principle, right? It's really useful to be able to like put in working hypotheses, research notes, reasoning, say what it's related to connect things up, like tracking questions. I know you are huge on this. Like mm-hmm is, is like one of the most important pieces of thinking you can do is constantly track. Whether you've answered them, whether you have good working hypotheses, like right.
00:31:10
How confident you are in those answers, that's all really valuable stuff. Yeah. Yeah. Totally. And like, like part of why I think it's so useful is one. I think questions are one of the most helpful organizational primitives of like anything in investigative work, you know, like, like I organize all of my claims and evidence, like in terms of what questions they're relevant to, you know especially when I.
00:31:36
Have formed those questions ahead of time, but like also, you know as your, as your own to workspace matures and you put more and more things into it, then you can start translating these questions into queries and pull up. Things that might have been helpful to you before you even thought of creating the question, you know?
00:32:02
Cause cause like the normal mode is, oh, you know, like I can tag, I can reference this question on all blocks. That might be helpful towards answering the question and that's useful. But like what about all this stuff before you came up with the question, just write a query that that will pull. Yeah. Yeah. So yeah, I even like curious to see how other people end up structuring their relationships to questions, cuz that's when I know I'm not
00:32:29
confident in my approach yet. Okay. So going into the last leg here, I guess, like what would you say you are most excited about for the future of, to, and tools for. I'm excited about this move towards ontology based apps. There's like a couple of them coming out, but I haven't seen any that nail the UX the way Tara has, which is why I love it and why I'm so like quite
00:32:58
certain, I'm gonna commit to it. I had, I had gone into this being like, I'm gonna do a scoped experiment for a month and see how Tara feels. And, and now I'm like, okay, I, I would stay in this system. Like this is great. And also the fact that it, I, I also have an escape patch. I did say I told the, the slack that was important to me. This download is J is very important to me. Yeah. I can get out of here if I need to, but like, I don't want to, so that's the good thing you're like not holding people hostage or letting
00:33:25
them voluntarily be part of your app. Of course I find that exciting. I'm definitely biased in that because I'm now at like an AI and ML company, bringing machine learning assistance into mm-hmm tools for thought I'm very, very bullish on. Obviously G P T three is still like in its NACY. We haven't really figured out how to use it well, or, or just like a generative transformers in general.
00:33:54
But bringing those into here and being able to interact with the system that can help me think help prompt me, help things like rewrite them. Mm-hmm just help me generate new kinds of thoughts. I wouldn't have been able to do on my own. I would, I'm excited to see that bit of the industry develop. Yeah. And I'm not quite sure how these. Tie together, like the established systems like this, but I don't think it's a far crier to no, bring something like that into something like you.
00:34:20
Yeah. My, my theory is like you have people that are exploring the extreme end of. You know, with AI, we don't need to require the user to do anything in structure because AI can handle it all. And I think that it's useful to have people exploring that extreme, but like, I believe that the correct outcome will be somewhere like the U give users
00:34:48
the ability to create structure and. And create data that is machine readable and then the AI's gonna be able to do a lot more, you know? And and, and, and I think thatt gives you really powerful tools for creating this structure, like on the fly out of nowhere, you know, and it's all in a tree. And that tree has items that has nodes that.
00:35:16
Fields and properties on it, you know? So like I'm excited to see that direction. And I'm also excited to see the illicit plugin that might or might not happen in the future. Yeah. Yeah. That would be great. That's all I really need in here. Like, I haven't found a great way to integrate academic articles as much yet. Mm-hmm but I'm kinda like waiting to see how that side of things plays out, cuz like something like a Satero plugin or elicit or both of those combined. Yeah. Like that would, that
00:35:41
would be. Yeah, totally. Yeah. And then what advice would you give to new users of, to, mm, it depends so much on their, what they want to do for it. I mean, my, my first piece of advice is like, make sure, you know, what you want from the system, like have, have a purpose. Like you are out outcome focused. It is easy for me to know how to structure this because I know what I want out of it.
00:36:11
It's like, I want to like write pieces that are well researched pieces of writing and track questions and problems that I'm working on at work. I think if you went in here without a clear definition of like, what kind of things you are producing from it or how it's going to make you help you be more I don't wanna use the word more productive to create better things. Then it might be a bit. Well, like how do you structure that? Great. So have a clear goal and then probably just start simple, like,
00:36:40
like make a couple tags, like don't think too much about it. Make a person tag, make a company tag. If those things matter to you, like don't try to make all your tags up front, like start putting things in and then like, figure out what tags you need as you go. Mm-hmm . . Yeah. Cause, cause clearly you can just create tags as you go too. Like, like as shown by the way that you just created tag for airlines and then suddenly you were able to create table out of, you know, your flight data, right?
00:37:09
Yeah. My favorite tag is the meme tag. I only have one name here so far, but I was so excited to make the meme. I like tweet tags. But I was like, yeah, I can now collect memes cause I have a meme tag. This is new, but nice. You can take literature notes on memes oh, nice. And, and that probably is appealing to the digital anthropologist in you, right? Yeah. Yeah. Yeah. I want like a whole beautiful card table of memes and their meetings in here at some point mm-hmm mm-hmm yeah.
End of transcript