gistfile1.txt

Hey everybody, it's Eric Torbert, co founder partner, village global, a network driven venture firm. And this is venture stories, a podcast covering topics related to tech business with world leading experts.

Everybody, we're now accepting applications for a network catalyst accelerator program founders in our program have gone on to raise money from Lux spark a16z slow first, round Susa homebrew mavar on obvious and effects signifier and many more, Learn more at apply at village global.vc slash network catalyst. One note I wanted to add this podcast is since recording it, we've made an investment in Rowan researches seed round and are delighted to have them in the village global portfolio started from the podcast now we're here. Hey everybody, welcome to another episode of venture stories by building

Global here today joined by a very special guest Connor white Sullivan, CEO and co founder of Rome research, Connor. Welcome to the podcast. Thanks for having me. So kind of by way of introduction, you know, I've been really excited to have you on the podcast for a while, why don't you sort of introduce Rome? And what is the insight that that led you to start it? What are you really trying to do achieve with it? So Rome is a tool for building personal knowledge graphs. So you could think of it like a way we described as it's like an outliner, or, you know, like Google Docs, except it's built on a graph database. So you don't need to think about Oh, what file or folder does my idea go into, especially if the idea like many ideas do relate to many different topics, it's really easy for you to to draw connections between different ideas to group ideas into sets of intersecting sets. And, and yeah, basically, it's a tool that is designed to work more like your brain naturally does and not sort of force you into artificial constraints that that the normal tools, we have for

knowledge for organizing knowledge. Do I want to ask about sort of the bigger philosophy behind Roman and sort of the the work that you did to to uncover that? And maybe one way to do that is like, what is to just dig deeper there? Like, how does our brain naturally work? And how is it How is it held back? Normally, when you have an idea, your your brain is drawing synaptic connections between all the other ideas that that thing is connected to. So if you are sitting on a beach reading a book about pandemics, you know, that was recommended to you by a certain friend, you're going to be able to remember the friend by that book, you're going to be able to remember the book by that friend, you're going to be able to, you know, think about Oh, it relates to things that are relevant for thinking about, you know, the COVID pandemic or something like that. So normally, you know, our brain can organize things hierarchically, but it also organize things associatively. And since, you know, Word documents and the file system have come out, we've tended to try and force our ideas into you know, a single page or

A page that gets into, you know, one file and a filing cabinet. And we've we don't have in the ways that we represent our text, the ways the the associative structure or the the sort of natural graph structure that that our thoughts tend to have. And so that really limits us both in, you know, remembering things or drawing the connections. But it also limits our ability to sort of do complicated reasoning of saying, Well, you know, this point implies these other four points. And if I invalidate that point, maybe I should go check up on those and see whether I still believe them. And you want it social, too, right. for about the last 12 years, I've been thinking a lot about collective intelligence and thinking about what you know, now people are talking about collaborative sense making and game B type things. But I originally got interested in the problem, looking at things like Wikipedia and open source and the way that you could have sort of large scale collaborative creation of intellectual comments, like the way that many people could collaborate to create a single information artifact that that could then be used by other people.

But there's problems with with Wikipedia, like, you know, it sort of tries to force a single point of view. And it doesn't. It's it, it grounds out with, you know, whether something has been published by the New York Times or by an existing academic journal, it's not a good place for original research. Because with original research, you need to be able to have some perspective, like, you know, one person is pushing forward a certain thought, I've been working on that space for about 12 years through, you know, one startup that and then after we got acquired through a r&d group, but we basically have thought that you need a different way of representing the structure of thought in order to enable sort of large scale collaborative research or collaborative knowledge creation. And so Rome is starting with that, that single player,

data structure almost, that allows you to collaborate with first your past self, right, synthesize ideas from all the different books you're reading and start to build these maps of the domain, even if that domain is

You know, your goals and your projects, or, you know, reflections over journaling or things that you've done over the course of your life. But to start with that, that single player tool, and then to enable eventually, much more collaborative thinking and collaborative problem solving, and flesh out what that looks like a little bit at scale, you just talked about that as a wedge, you've mentioned you use Excel as a Northstar. But in a world in which Rome, you know, achieves its dreams, how is the world different? Or what does that look like? Well, so one way to think about it is just the, the way people organize book notes right now. So I was really inspired by a book called How to Read a book by Mortimer Adler, where the idea is, if you want to form your own thoughts on a domain, one thing you do is you read, you know, 20 different authors on that domain. And you break those authors down into what questions are they asking what are the claims that they're making, what's the evidence they have for those claims? What's the structure of the argument that they're presenting

What kind of key terms are they defining? And then you map those ideas together. So you say, you know, how does how does this one author support or oppose one other author, you start to actually synthesize those things together. And, you know, before I started Rome, I ran a bunch of collaborative research groups where we would read different books on the same topic. And we try to organize them in tools like you know, workflow, your Evernote, like Dropbox paper, but because you couldn't have all the you couldn't easily remix these notes, especially if they were originally organized by by book, right, you can't suddenly see all the questions from all the authors or all of the quotes from the relate to one particular topic that you're you're looking at across 20 different books. So yeah, what what it looks like is it looks like the ability to to have many different indexes into your knowledge that are really easy for you to, you know, in Rome, they get generated somewhat automatically through this thing called bidirectional linking where

You know, when you type a word out in and wrap it in double square brackets, it automatically creates a link to that page. And it'll also pull in that paragraph, and anything that was indented underneath it into that page into a collection of backlinks. So it makes it really easy for you to have many different indexes into your thoughts and different ways of sorting them and filtering them. And then, and then pulling those notes up into synthesis documents. It is interesting. So you know, we talked about sort of the shortages of Wikipedia, and you've talked about threads like Twitter, to some people, you serve some elements of this construct of, you know, closest or to me personally, it's the closest thing I use to sort of a global brain or collective intelligence vehicle, ie I sort of think out loud in it. It's sort of use it as sort of a collaborative Google Doc, I get feedback on various different sort of, you know, tweets, if feedback in the realm of likes, but also in the realm of actual like, you know, here's the right answer type of feedback. And then I try to create sort of this threaded structure and then threads and

Read structure and it seems like Rome is that you know, 10 x or 100 X. Yeah. So I think I think there's a, there are a couple of different ways to use Twitter. And I think they're one of the things we've noticed over the last couple years is that there's a small minority of Twitter users who are using Twitter. As an I'll quote, this guy visa Canvey, where he says he uses Twitter as Evernote with a slot machine machine attached to it. And because you never know, you know, some people will just, like scream at you for something you say, or they'll they'll post some some general insight. And Twitter isn't really designed. You know, we've talked to a bunch of people at Twitter and, and as a company, they're focused on now, right? They're focused on like Twitter as a replacement for looking at your window. And it's it's very present oriented. But some folks like yourself can sort of hack Twitter and they use search extensively. They search their old tweets, there, they're copying, you know, links from their old tweets, and they're, they're building these threads and they're, they're building threads.

Have threads and you know, they're they're re mixing old ideas into new contexts. And it's it's a pretty amazing and powerful tool because you know, when somebody starts following you and discovers you they can go down any kind of rabbit hole based on their interest and find, you know, tweets that you've written two years ago and and you know, replied to them then write reply to them now and and give you some new idea. So, yeah, I think I think the people who are sort of hacking Twitter in this way where they're, they're writing a book summaries like you do, right, or they're, they're forming and then from this book summaries, they're forming theses on the space and you know, re mixing those Book Summaries into or like the individual claims or the individual observations made in a book summary in there, they're pulling them into larger threads. It's a pretty amazing and powerful space, but it it also, you know, everything you write has to be in public so it doesn't leave as much room for you know, the thoughts that are more theoretical or you know, that you're you're not quite ready to to put out anything

Be open. And search is also really poor, right? It's very hard to find, you know, you'll see a tweet and like it, but if you don't immediately thread it, it can be really hard to find those in the future. So, so yeah, so the idea for Rome is like, and Rome right now is I would say, actually, like half of my personal knowledge management system, the other half being Twitter, right, where, you know, I, I have a, similarly a lot of threads of threads, and I can pretty easily find certain things that I'm looking for by remembering one thread, or like one tweet in one of those threads and, you know, following my own rabbit holes to find the thing that I'm looking for, but yeah, so Rome is Roma's taking that but amping it up for you know, notes that that are private. And yeah, eventually we will be moving into the more public social domain. Right. And so for improved, and improved searchability ability. Yeah, yeah. And I think the key thing is like there's a there's a difference between searching and browsing right. Searching

No head of time, what you're actually looking for. One of the things that people like about Rome is, so every every page in Rome is both a place to write and a query over all of your other notes for other places where you use that term. And so and there's there's two parts of that query, there's the link to references that you've formally connected to that page. And there's unlinked references, which is just, you know, places where you've used that term. So if I said something like, collective intelligence, even if I hadn't formally linked it in the past, I can see all the notes where I've written about collective intelligence. And so this, this means that it's easy for you to like, roam or like wander through your old notes to find nuggets that you want to reuse or that you want to put into a new context. And that's, that's sort of one of our core ideas is you should be able to get compound interest from your past writing. A lot of times people take a note or they save some, some bookmark, and they never revisit it again. And that really, yeah, the future

purpose in many ways, right? The goal for us is can you get compound interest from your notes? Can you, you know, can can each thing that you write be a knowledge artifact that is giving you more and more value over time by and when you can start connecting ideas together, that's when you really start to see the the power of the notes you've taken in the past. So we've seen people say, you know, they made 10 times more in Rome than they did a notion, and part of it is that they're able to see more continuous value from from those thoughts that they're, they're capturing. Yeah, a compound notes. I love that. And it's similar what tactically is, is my experience be within Rome, different than it would be on a notion that would give me that sort of TEDx, you know, or compounded experience? Well, the first thing is that, so we, we encourage people to take daily notes. So in terms of, you know, quick capture, you shouldn't have to think about, you know, does this relate to, you know, the books that I'm reading, does this relate to the company that I'm running, does this relate to

You know, hiring or like talent development or whatever, right? You just, if you have a thought you put it on the note for that day, and you reference through linking, which is functionally the same thing as tagging all the topics that you think might be relevant, and you sort of indent those underneath. And so your first sort of entry into this knowledge graph is just a quick capture of thoughts as you happen. So it's, it's, you don't need to think about the structure of your thoughts immediately as you're having them, you just sort of can brain dump and write. And so yeah, the folks who tend to get in the room sort of use it as like a daily journal, where they're logging, you know, their tasks, the observations, they're having notes that they're having for meetings, just, you know, quotes from things that they're reading that kind of stuff. But then, you know, over time, you can both observe a sort of bottom up emergence of the relationship between your thoughts and you can start to more formally structure it because every every paragraph or every you know, bullet point in

You can choose whether you view it as an indented list or you can turn off the bullets and just view it like a document. But every bullet point room has a unique ID. And so it can be embedded in many different pages. And it's also going to appear in that query section for, you know, a number of any any other references you make. to other pages, it'll appear in the no query section for those pages. Talk about this concept

of Zettel caston. Zero caston is a system that was developed by a German sociologist, he sort of, you know, invented systems theory in in Germany. And he wrote over the course of his career 70 full, pretty dense academic textbooks and, like over 400, like peer reviewed research papers, and he did this without any research team, you know, he said, you know, his research team was his passion, which was, it's German for slip box. So he had I think, by the end of his career he had like, over 10,000

And index cards, each of those had a ID, you know, a unique reference. And he would link those IDs together. So his practice was he'd write three or four of these notes per day. So it's very similar to actually the way I've seen you use Twitter, where, as he was reading a book, he would write sort of a summary of the book, in his own words, and, you know, he'd thread those ideas together. But then he'd also be able to, to go back and index those ideas into other theories he was developing or the questions that he was interested in answering. So

the, the, the mark of a good knowledge management system is it's able to surprise you, and it's able to be sort of like a good conversational partner, because, you know, it's, it's able to take, it's reminding you of things that maybe you'd thought before but you you haven't thought of in a while, and they mean something different in the new context of what you're thinking about. So, we think is, is the you know, it's sort of like

Get GTD for knowledge management. Because, you know, you can't put a deadline on insight, you don't know when you're going to have the breakthrough or when you're going to realize you're asking the wrong question. Or when you find some piece of evidence that invalidates the entire hypotheses you were working on before, you can't you can't plan for those things, but you can have sort of deliberate practice and like consistent habits that allow you to get insight faster, right. And by the time one of the reasons he was able to write so many papers and so many books is when it came time for him to, you know, produce a an artifact, right? Like a presentation, you know, class notes, like a lecture or or these books or research papers. He wasn't ever starting with a blank page, because he would, you know, grab one of those cards, then find all the cards that were connected to it, pull them out, and then he could shuffle them around on the table, figure out what the right structure was. And then you know, the paper, all he had to write was was the glue to tie the different ideas together. So by having that, you know, writing three or four index cards per day

He was able to eventually be really prolific and and produce a lot of, you know presentable materials out there, folks. And he is, is Dave Allen to get things done? Or who is this again? Oh, so this is this is Nicholas lumen, the two people who who followed that, that I actually discovered it from or Robert Greene, who wrote the 48 Laws of Power, I think was one of the first people in the English speaking world to adopt that system. And, and then his protege, Ryan Holliday, who, you know, wrote the obstacles away, and, you know, trust me, I'm lying, those kind of things. That's actually how I originally discovered this little cast. And it was, they had similarly, you know, boxes of index cards, where, you know, as they were reading history, they were, you know, making these discrete observations and then starting to connect those together so that when it came time for them to see some patterns across history, they already had the raw material to write those books that there were no Robert pirsig, who wrote Zen and the Art of motorcycle maintenance, and then his book, Lyla.

He in the book, Lyla, which is much less popular, the both? Yeah, he describes himself, you know, sort of rearranging note cards, but he also describes sort of the, like state of mind you need to be in to be able to be open to insight in the first place. Yeah. Which is, which is pretty interesting. Yeah. The way I'm a big fan of the idea that, that, you know, we don't really have novel and the, the insights sort of emerge through us, right, like, like, you know, Newton and, and leaving, it's sort of, or supposedly invented calculus almost exactly the same time, in different ways. So there's a sense that there's a, there's an adjacent possible of like, you know, what idea is ready to come into the world? And if you are, you know, if you're open to it, if you're making a bunch of observations and connecting them and you sort of are viewing the world with a particular question in mind, then you're you're more ready to give birth to those ideas. But if you don't have systems that allow you to, to draw new

connections. It's hard for those ideas to sort of come into existence. So, yeah. And why is it Excel for text instead of Google Docs? Well, so the thing that we think is really interesting about Excel is one there's Excel is actually a really, you know, it's a purely functional reactive coding environment, right, it actually gives you a huge amount of power to see how data moves over time. We think it's a really powerful tool, and there's 750 million people who use it, but many of the people who use it just use it to draw tables, right? They're there. They're not using all the crazy functionality that's available for it. And so with Rome, so we like the idea that you know, a tool should be really easy for people to get started with, you know, it should be as simple as just writing lists out. But as you have more pressing problems and you learn to use the tool more you should be able to, you should be able to get more power out of school. So we talked about sort of a low floor and a high ceiling, like it should be

To enter, but then as you learn to use the tool more and develop more strategies for using it, the tool should have the capability for you to get more power out of it. And so that that's something that Excel has, right where it's very easy for someone to get started with it, they don't need to learn about, you know, all the crazy functions that are available to it. And they don't need to be using pivot tables, you know, when they first start. But as they learn to use the tool more they can you can eventually you can run whole businesses on Excel, right? There's there are many, many, many businesses that are just run on on Excel tables. And so that's kind of the the model that we have is it should be easy to start with, but then you should basically be able to code in the tool itself and be able to do algorithms over your previous thoughts in order to generate new insight. Go deeper, a little bit into I'm curious where it's sort of the how you're sort of you've been looking at this for 1213 years, how is your sort of understanding of knowledge management, or sort of collective intelligence evolved in at a time that you

There's any sort of sharp changes in how you thought about it, or how it evolved along with the as the field has evolved. So, I've been really influenced by Michael Nielsen's work, he has a book called reinventing discovery where he talks about, you know, what network science would look like. So how could you and, and a good example there is, you know, if you have 100 people in a room, you know, 100 scientists in a room, if you're doing everything sort of synchronously, you're limited in terms of how much intelligence you can get out of that whole community, adding, you know, someone might end up spending, being sort of bored in one conversation, they're not necessarily getting matched to the right, other people in the room to talk to, you know, there's, they might be slightly tired, right? There's, there's, there's challenges to synchronous collaboration in a knowledge space. But if you're limited to just the way that we structured text right now, where, you know, 100 people are in just a discussion thread or like, you know, a Facebook Messenger chat, right? It's the same sort of problem of you

can really easily get overloaded with with too much information you have to figure out okay, like Where, where, where can I deliver the most value to, you know, helping this group solve this problem, it's hard to navigate that if the idea is don't have any sort of structure inside them. So my, my first startup was trying to solve this in like the political domain. It was a startup called the block cracy. And we were trying to make a online town Commons sort of vision was to like wicked by government, and to basically crowdsource ideas from people who were registered voters in the town. So we can evolve, you know, solve some astroturfing problems, and have them, you know, vote up proposals and, and sort of vote up to pros and cons. But there's, there's a bunch of problems if you have, there's often a low trust, you know, rivalry dynamic in politics, where, you know, people are incentivized if, if you're joining a platform, and you know, the platform currently is mostly filled with people who disagree with you and they've raised some really good points. You could just avoid using that to try to discredit the platform itself. And and say like,

You know, okay, that I'm not even going to address the the points that that other people have made, because I can, you know, I can just push the whole project aside. So at first at first, I was really starting with that, that collective collective intelligence thing. And we realized, you know, one, a tool should be useful first for the single player perspective, because if, if you need network effects to get started, it's just it takes a lot longer to to Yeah, to end especially if you think that the the real insight comes from the intersection of totally different domains. If we want to have you know, molecular biologists collaborating with Game Theorists collaborating with, you know, physicists, you shouldn't need all the Game Theorists to be on a platform for it to be useful. So one of the ways that my thoughts have evolved is we sort of shifted, even though we have these sort of big visions for collective intelligence we, we've shifted to just focusing on a single player tool first. So that you know, you don't need to worry about some critical mass in an idea space.

So yeah, so that's, that's the first place. You know, in the course of my research, I spent a while thinking about formal logic and predicate calculus and like, you know, figuring out how you sort of, like map mathematically map out a logical argument and make it so that, you know, if I, if I change the truth value of one thing, it'll flow through the whole system, but I think that we sort of realized, actually, like, regular human language is incredibly powerful. So if we can just have, you know, pointers to different ideas, and, and lean on, you know, the logic of human language that actually is, is a lot more useful for a lot more people right off the bat. So we're still interested in you know, figuring out how to, you know, there's there's calculations in Rome, there's the ability to sort of have numbers propagate through through a system but we've been a lot less interested in, in the sort of, like, Bayesian reasoning, like mathematical stuff, compared with just how do people define terms and and, you know, point to a definition and

You know, like, also just just take notes on on whatever and not worry about formal formal reasoning, but, but just drawing connections between ideas that they hadn't drawn connections to before. Have you gone deep on sort of linguistics or how language works? I've been really influenced by Douglas Hofstadter, who wrote girdle, Escher, Bach. And then his his most recent book, which is well summarized by the lecture he gave at Stanford, called analogies, the core of cognition. And so he talks about analogy is pretty fundamental to human reasoning. And I think that that's one of the places that that Yeah, I've been, I've been most interested in terms of like, how to model thought it's like, how do we model analogies and how do you how do you connect, you know, concrete examples to an analogous thing, right? Or analogies abstraction over those thoughts.

How do you envision Rome interfacing with sort of new platform shifts eventually whether it's like VR AR

You know, HCI stuff. There's an exercise that I did a couple years ago where the the prompt was to sort of imagine a day in your life, you know, 10 years from now. And like, what is that day look like and, and for me the day sort of started with like, climbing up a mountain with like, ar glasses and being able to see a full like, as I'm, as I'm hiking, being able to manipulate, you know, my thesis on the world and the structure of my business and, and being able to actually, like, see in sort of three dimensions, the relationship between a bunch of things. So I think there's a lot of, there's a lot of opportunity in AR like I'm a big fan of Brett Victor and and, you know, I think that if we're like limited to just screens and you know, two dimensional manipulation of ideas, it's harder for us to see the matrix and sort of the, you know, yet the, the the many ways that that thoughts are connected or that the problems that we're working on are connected. So I think being able to move in

To three dimensions or, and more is important. It's just you know, it's I've been really influenced by Douglas Engelbart, who's like most known as the inventor of the mouse, but in many ways is sort of the father of human computer interaction. Back in, you know, the 60s and 70s, everyone was trying to automate everything away, right, the focus was on artificial intelligence and figuring out how you could replace, you know, replace humans for operations. And, and he sort of had this idea that, that what you should be doing is giving humans more power, and giving them more capability and augmenting intelligence instead of focusing on AI instead of, or intelligence augmentation as opposed to artificial intelligence. And so, but one of the ideas he talked about was bootstrapping of, you know, you should be building tools that augment your intelligence and then using those tools to solve harder problems. This is this is also the sort of flywheel that folks talk about. And so, you know, I think right now

Now our goal is just let's build the best thing that we can with text and with two dimensional representations of, of, you know, knowledge relationships. And then when AR is ready when you know that the

that the landscape is there that it's easy for us to build on top of those platforms, then I think there's a lot more capability that will come out of that. I had this this tweet last year is that people are often reading, curating, discussing the same books, articles, podcasts, and so frustrating that this work doesn't seamlessly build on top of each other, that we don't really benefit. It's a structural problem. And then as that annotations transcriptions wikis, we'll get there is this what your hopes are hoping multiplayer evolves into, you know, blog, fascinated with things like, you know, Rap Genius, who was our Wikipedia earlier, like, yeah, are you sufficiently that? Yeah, I think there's, there's something that is, it's pretty magical when you find somebody else who is reading to obscure things in two different fields, and they're connecting them the same way.

That you are, right. So that's, that's one of the things where I've always been really interested in being able to pivot off of content to, you know, a social network of people who are interested in that content, pivoting off of that into other content that they're interested in as well. So one of the things that that, you know, we care about is how can you, if you are, you know, annotating some book, or, you know, like taking notes on on some YouTube video or on some podcast, and you start to connect ideas in, you know, like, so first is being able to see the other people who are also interested in that, that same thought object, whether it's, you know, a book or a podcast or a video, but it's more interesting when you can find other people who are connecting it to something else, that's niche, because you're, you're likely to be to have a pretty strong affinity for them. So being able to do that sort of form those ad hoc research communities.

And then that doesn't just apply for folks at the sort of cutting edge of the field. That's, that's where

we've we've been sort of focused for now. But it also is, you know, if you're a 12 year old learning calculus, and you're, you know, but you're really interested in being able to launch bottle rockets, right? How do you find other 12 year olds that are at around the same level of competency that you are in the same level of interest you are and find the other material that has worked for them, to help them learn solve the problem or move closer to the goal that you have? So yeah, we're really interested in these kind of like ad hoc social research communities, very broadly defined, totally the way I sort of got into this construct of so go one at a micro level, you just been fascinated by these communities and wanted to participate in them and just sort of, you know, naturally using Twitter the way that I do, and then it sort of sort of a theoretical macro level, you know, I think nonzero by Robert Wright was the book that got me into it. Yeah, describes how, you know, the story of history can be told by sort of an ever increasing sort of social complexity and in

intertwining of our fates, yep. And that, you know, we become more and more interdependent? Yep. Yep. Well, that's, that's, that's the thing that's cool about Wikipedia and about open sources is you have almost a rivalry dynamic where someone is someone wants to be the Wikipedia that that, you know, like, has made the last edit on that page, right? And you can have some some social competition where where people are trying to like write the best Wikipedia article.

But the, the result of that sort of rivalry dynamic is something that's totally non rivalrous, right? It's a, you know, an object that then gets used by by everybody else. They can be fighting for sort of social status in a very limited domain, but then be able to create something that creates abundance for for all the other people who are watching. And so yeah, I'm curious though about like, how you started using Twitter the way that you're using it, like cuz there aren't, there aren't that many people who are

Thread quite the way that I've seen you threading. Yeah, it's a good question I, I used to in a few different ways. One One is ask a lot of questions or sort of the Cunningham's law, like, get out? Yeah, lots you'll get, you know, put every response might be wrong might be simple, but and someone will make it smarter. And also just the feedback loop of, I'm now sort of not addicted to it. But I definitely felt the buzz of like, Oh, I like getting feedback on Twitter. Thus, I should get more interesting ideas that I should be more, you know, that's Yeah, honest feedback loop for me. But in terms of threading part of it, I saw other people doing it. And then sort of I realized that that would be a good way for me to sort of crystallize or codify these, these concepts. Because Yeah, as I put them in sort of natural orders build on top of them. I'm not an engineer, but I've always been excited by the idea of like GitHub for for thinking. Yep, yep. Yep. Actually strive to do is take some of them and put them into a monkey. So you remember that? Yeah, they get up for thinking. More cutting him, the guy behind cutting him saw also

The original author of the wiki, his latest project, which he's been working on now for, I think, over a decade is called federated wiki. And the thing that really attracted me to that was it had a model more like GitHub, and that's, that's sort of how we think about collaboration is, you know, every, every person had their own, you know, their own wiki, but and it could be connected to certain networks. So if someone wants to make a change to one of your pages, they make a copy into their database, and they make those changes and they're basically issuing pull requests back to to, you know, your wiki, which you can accept or reject, right. And so I think that that model of every person, there can be gradual consensus, or, like, you know, gradual emergence of different consensus is a pretty powerful thing in GitHub, where like, each person, you know, has their own copy, they can make a branch they can, you know, then ask to change back into, into master on the source. And I think that's, that's a much more powerful model for collective intelligence.

You know, because you need to be able to just like when you have a Google Doc that has 100 people editing it, it's, it's pretty stressful for you to, you know, make major structural changes to that document. Because you know, you don't know how people are going to feel about it. But making those structural changes is how you, you know, have the new idea or see things in any way. So you should be able to just make your own version where you, where you make those changes, instead of just copying and pasting and losing all the connection to the original source, which is what people do right now is they'll just go, if they want to make a major revision, they'll just make a new document and all of the connection to the source gets gets totally wiped away. So yeah, GitHub for for knowledge is is another way of thinking about what we're trying to do when we're Yeah, I love that I've always said I want something that has sort of the sort of searchability or sorbo and maybe you guys are just 10 x better of Korra have combined with sort of the real time this is not even real time this but just know the feedback and engagement of Twitter and again

loops, because core has such good content, yet I'd never checked it tutors contacts kind of, okay, your hitter has a check every second. I mean, in Twitter, Twitter has this graph structure, it's just, it just doesn't expose any of the affordances in the UI, right? So if you're looking at someone's thread, there's no way for you to easily see that, you know, three of those threads have now been, quote, tweeted and put into a new context by somebody else, or by the same person, right? So you're only able to sort of see these links pointing in one direction, which makes it, you know, much harder for people to go from an individual threads they're seeing to seeing like, Okay, how is that? How's that thread been recontextualize or, or synthesized into a higher level idea. It's easy to drill down, it's harder to zoom back up. I'm curious. Are there any debates in the space or dessert different approaches to to knowledge management that you're interested in or have opinions on what's your thread I saw that Venkat said that you and Tiago

had a different approach. Yeah. So, I'd like a lot of tacos ideas, Tiago forte, one of the main ideas that I really liked from him is this idea of intermediate packets, which I think just fits, that'll cast in perfectly that, you know, if you have a big project, if you can break that project down into smaller concrete deliverables that could be potentially reused for other projects, it's easier for you to you don't you don't have as many sort of hard stopping points where the whole project has been has been worthless, right? Even if you decide that the blog post that you're writing isn't actually that good. If you've had these concrete deliverables of, you know, smaller observations that that you've written in your own words and you phrase nicely, those might be reused for for many different things in the future. So I like his idea of intermediate deliverables, but Tiago is, as I said, like he said, he said some ridiculous stuff like you know, all the early visionaries ideas were fully realized by Microsoft Word, right and that you

No, he sort of has this idea that the tools don't matter, that, you know, all that matters is the processes you're going through which, which sort of makes sense, because he's a coach that like, just teaches process and wants to reach largest audience. So, you know, he tries to be tool agnostic. But the thing about tools is that they, they don't, don't just map to the tasks you have, but they also have affordances, for for thinking about new tasks that you might not have, you might not have realized before. So like, you know, you can climb a tree with with like climbing spurs, and it might make it better for you to climb a tree. But if your goal is to pick some fruit, you know, having a ladder totally changes your relationship to that activity. Or like, you know, excel at first, you're just using it for calculations, but it also makes it easier for you to do sort of like as if comparisons or like hypothetical comparisons and and so you end up having a whole new set of workflows that can emerge because you have the capabilities of the tool. So, you know, obviously, I believe in building tools, because, you know, we shape our tools, and thereafter our tools shape us and so I think that I'm a bit

Technical determinists where I think that, you know, there, there are certain patterns of thought that get baked into the tools we have, and that there's a problem with the way Evernote structures it or the way, you know, notion structures. You know, they're their notes where everything has to fit into one precise bucket. And it's very hard to remix things. And it's very hard to, you know, look at the intersection of two ideas and see all the notes you have about both of those ideas. And so I I care a lot about processes that emerge for this, this new kind of tool where you're, you're thinking about knowledge, graphs, and you're thinking about re mixing and sort of a first class object. Yeah, and that's sort of the big, you know, beefs that Tiago and I have are, you know, I think he's done a there's there's a few ways that you can augment intelligence there's,

you can have artifacts that that change the way that you relate to your tasks, you can have methods and sort of practices like, like a Spaced Repetition habit, right. And then you can have language like the language of intermediate packets, right?

Or like the language of game B or something like this, like where the the language sort of causes you to focus your attention on a different part of the problem, right? Like languages is incredibly powerful tool for taking, when you make a word for something that previously was only described in like a full sentence or a paragraph, you know, now that term can be composed to have sort of higher, higher order thoughts. And I think I just think the tools actually matter really deeply, right? Like the tools that you use really do shape the way that you think in a lot of ways, and that I've done hundreds of user interviews and the level of stress that people have had with other notetaking systems, right, where they're like, they're, they're stressed about the fact that they know they're going to lose the note. They know they're not gonna be able to find it in the future. They don't know what to title the note or what notebook to put it in. Right. I think that a lot of the insert. There's a lot of incidental complexity when you try to use this strictly hierarchical knowledge management system, and it doesn't lend itself to seeing a sort of bottom up emergence of higher

Your thoughts are naturally connected. Staying on tools and tools are important. What else is in your sort of requests for tools outside of Rome, like if you were, you know, had a year long hackathon for other people who said they were gonna work on things separate from Rome, and I was there, I was gonna build a lot of this. But the there were tools that were gonna move us forward in some way. What else comes to mind? I think I, I particularly care about, like algorithms of thought, right? So when I think about, like, Mortimer Adler had this protocol for how to read a book, where, you know, you would you would systematically skim the book, you'd look at sort of the first and last chapter of the book to map out the argument, like, and it was, it's sort of an algorithm of allowing individuals to, to parse to parse it a data structure that's basically formatted as a string into something more like a like a graph. And I've seen a bunch of other these kinds of, you know, like Jordan Peterson has that that site, which is, you know, self authoring, which is basically just giving you a set of questions, the answers to your questions become inputs, then

next set of questions. And it's a very simple, like algorithm for helping people define what they want out of their life and sets and vision. So we've been working on, like, basically building out a sort of third party ecosystem for people to, to make visualizations over, like, data structures within their notes. So you can sort of map over your thoughts and sort of ask a follow up question for each of those or you know, map over a set of

a set of constraints and a set of options and figure out like, you know, what, what idea fits those best so I I'm very bullish on just general, intelligence, augmentation and ways for people to think about patterns of thought and then, like, you know, unique you eyes for, for basically, yeah, running running algorithms where you are the compiler, you know, where, like, where, you know, each each step in the algorithm is causing you to have a new idea and write something new or draw a new connection. And I think there's just a ton of like, like, the space is really untouched, so right

So there's a lot of you guys that haven't been invented, like, you know, the closest I'm thinking about is, like type form, lets you put like logical conditions into, you know, if somebody answers the question this way, this is the next question we get them brought to, but they're still haven't been us developed around, you know, taking the answers to some of those questions and putting those inputs into into follow up questions. So, I'm pretty I'm pretty bullish on on that. And I think that's, like not just useful for the single player mode of getting you in conversation with yourself to like to have more insight, but I think it matters a lot for if you're trying to do any sort of collaborative knowledge work with a large people. So But yeah, I mean, I also hope that there's just a lot more progress in in AR and you know, better like better inputs for like audio to text and those kind of things. What do you think about Andy Matuschak, and Michael Neal's mentioned my goodness earlier about the work that they're doing now, and where does that sort of intersect or differ from

Does that work better? Yeah, we, I talked with Andy all the time. I'm actually I've got a conversation with Andy later today, the work they're doing on tools for thought, right is there's not a lot of people in that space like there and they're doing really, really good work. And so I've gotten a lot of ideas from Andy, I've been happy to see some ideas from Rome incorporated into you know, like, like, contextual backlinks. Is is a thing that you can see in his public notes. You know, he baked that into his, his personal knowledge management software. So I hope that the ideas are, it seems like it is or going both directions, his ideas around programmable attention of basically saying that, you know, space repetition, which which, you know, they've popularized with with quantum country and with Michaels stuff on on key is actually a, you know, it's about, it's about managing your attention allocation. And it's not just useful for memorization, it's useful for you know,

rephrasing an idea over time or prompting yourself with some daily question and then answering it a different way each day. And so we've a lot of our ideas for the next year, the things on our roadmap actually relate to some problems that Andy has articulated really well. And I think he's, he's coined some terms that have been really useful for us like programmable attention. Image Ravel was before and we're going to be talking about what you think are the biggest contributions of of game be I guess now we're moving more at a higher level. But where do you see that going? Or what's most interesting about about what, what they've got sort of going on there? Yeah, I mean, so one of the things that I've been trying to figure out, like when we eventually get to a place where people are able to build knowledge on top of other people's knowledge really easily, right? instead of like copy pasting, right, they're able to just directly embed content from other people.

I've been thinking about how do you both incentivize people for building these like atomic units to get reuse like any

You know, yeah, like, what are the incentive structures look like for,

for contributing to the commons? And how do you also basically allow people to make a living by doing knowledge work and, and, and creating really useful, elegant explanations of complex ideas or like ways of teaching people skills and those kinds of things. So the thing that I've been really interested in, in game B's, the problem of incentive design and the problem of value capture versus value creation, because I think there's still a bunch of unanswered questions that are still unanswered for us, in terms of, you know, how you do incentive structures when you're trying to create an Information Commons. And, you know, basically, yeah, like, how do you how do you both do incentives? How do you do incentive structures and how do you also make sure that people are able to make a living who are creating a lot of value, like people are able to support themselves and so you

There's mostly good questions in that field that haven't yet seen great answers. But I've also, I've liked the phrasing because originally I'd been interested in this this problem of, of the commons coming from Yokai Benkler his work where he was, you know, he had some great papers in like the early 2000s. About Information Commons and copyright and, you know, yeah.

And it's nice to see that in a more in a less academic language and, and, and getting more popular across more people. So, yeah, that's kind of that's how I've been thinking about game B is it's about getting people to ask the right questions. And, and thinking of that institution designed to but but, yeah, yeah, just the value creation. Value capture is really interesting in the sense of if we want people to have you go back to Michael Nielsen, if we want people to have you know, make more observe breakthroughs and, you know, science, you know, that will then lead to all

This value, you know, creation or being captured by other people who they don't see, you know, the Eric Weinstein always says, hey, how come you know physicists actually see the, you know, the the value that was created from, you know, people who leveraged it to create other things? How do we best incentivize people to do that? Yeah. Yeah, I was, I was actually just about to bring up Eric Weinstein, because, you know, his sort of thesis that like, all of the progress of the 20th century was, was riding on the back of physics. But, you know, physics doesn't, like the only way you can get a job in physics right now. I mean, some folks end up going into into hedge funds and, you know, being quants. But but there's, there's not a great there's not a great way for any physicist capturing the the value that they're creating.

And so, yeah, our society doesn't, like, allocate nearly enough resources into fundamental research and, and and even the, like, clear articulation of that fundamental research, right, like writing a really great you know,

book that popularizes physics like, it's it's nothing compared to the value that's actually being created by those by those concepts. But Gladwell has it sort of the idea of basically redefining private property, but then we get into the realm of, you know, it's just hard to think about. Yeah. Yeah. I mean, the problem that we've been interested in thinking about is like, how does how does a small group like, you know, our communal house make decisions? Right? How do you figure out how to do these problems of, you know, attention allocation and gradual consensus, like how do you do these amongst small groups of people where there's already high trust? And can we do them in ways that like, that, that can scale? But it's, it's still a, like, it's a big, there's a lot of design problems around, you know, attention allocation and and, like, figuring out what, what you bubble up to somebody at what level of granularity and how can they like, keep zooming in.

And, yeah, have those contributions? Yeah. Especially if there's a logical structure to what's going on, like have something that

Is it zoomed in at a very fine level? If you disprove a certain axiom have that still, like bubble up all the way up the chain so that, you know, you can when you're reading a paper, if the paper, you know, cites something to fail to replicate, like get sort of a warning as you're reading the paper that like, hey, the whole premise of this, this idea might have been invalidated? Like you should, you should check the work. Yeah, totally nearing the end here. Is there anything we didn't get to cover that you think might be interesting for our audience to know or perhaps, plug anything else about Rome in terms of what's what's what's to come? So right now, we're, we're in a place where a lot of people's calendars have changed a lot, right? They're there, they're not going to be out socializing as much, you know, there were, at least for you know, the coming months, we're going to be in in a little bit more of a quarantine or social distancing space and I think that the there's a ton of opportunity there, right?

You know, Newton invented calculus, Shakespeare wrote King Lear while they were in quarantine. And it's a great time for deep work. And it's a great time for trying to clearly articulate like what you believe. Like what theses you think about the world, and what questions are really interesting to you. And so, I've, I've been,

I, I'm pretty hopeful that some really cool things could come out of this time. And obviously, we think that Rome is is the best tool people can have for, you know, generating insight for making progress on the hard problems that they've not been able to make progress on before.

So I'd highly recommend people read some Aaron's book, how to take smart notes, which I think introduces the the idea of Zettel caston and the practices around, you know, when you when you read a book, like, you know, copying and pasting or directly, you know, directly quoting it isn't nearly as useful as articulating it in your own words. And then what's most important is after you've articulated in your own words, figuring out how to

I get connects to other things you're thinking. So I'm pretty hopeful about, you know, this time is I think it's, it's, it's a great time for Rome users. And so, yeah, I hope that folks who are listening your podcast check it out. I would also say, if I was just going to make a quick plug. So Rome is Rome is a power tool. Right. And, and it we have not, we're still in our very, you know, earliest stages. Our first couple years, we were actually just funded by the AI safety in sort of existential risk research community where we were, we were getting paid to develop the tool for for a very small group of people.

We still haven't mastered onboarding. So it's, you know, we're leaning on, there's a huge community of people who've been making YouTube videos. And there's some paid courses out there for how to learn to use Rome really effectively. So if I was just, you know, pitching, pitching the tool, I would say, look at the community, like

Get on, get on Twitter and start, you know, at least using Twitter as is that'll pass and but

but if you're, you know, look at the the room called hashtag. And if you're interested in room, like, take a look make sure you search YouTube and and search twitter and take a look at some of these courses. I love that. That's perfectly throughout my guess it has been Connor Oh, I told him Rome research. Connor, thank you so much for coming on the podcast. Thanks so much, Eric. Appreciate it.

If you're an early stage entrepreneur, we'd love to hear from you. Check us out at village global.vc