 These are office hours on Thursday, July 7, 2022. There's a few things going on in the world that are distracting. That is too bad. And I just wanted to set some context here also the the relate wiki website is behaving badly, a thing that Pete fixed. Oh, wait, let me just do. Oh, I just got rid of a tab that I didn't want to get rid of. So a thing that Pete fixed just a couple days ago broke again so I don't really know what's happening but I can use obsidian over here to just talk it through and screen share that will work. And earlier today, I was on a call where I was trying to use the wiki to talk about stuff and I couldn't find what I'd done and I got, I got discombobulated which was not not that encouraging but I think it was just me I'm not having my brain set right. Any thoughts or questions or anything like you guys have almost no context on what these office hours are what relate is and so forth so happy to answer questions, or I can just dive in and try to make the context. I'm glad you brought up the, the relate wiki. I was only if it was just me. No, it's, it's not you. It's yeah. And it's not me. And then, I thought I'd stop in we could definitely talk about the mid range goals which would you have on the task today but also I was having a little trouble anchoring relate in my brain. So I wanted to spend a few minutes finding the, you know, kind of like when you're trying to figure out, like let's say you're putting the other puzzle and don't know what the picture is sometimes you kind of make the borders first. And so I just want to kind of see where that vision kind of stops or if it doesn't, which is okay. Perfect. So I can do that. Bill anything you want to ask read. No, I just Bentley just did the, you know, the general systems thinking perfect intro draw the boundaries first draw the boundaries first thank you now. First you find the pieces that have a straight line on them and then especially the ones that look like a corner and you put those at the corners and then you work you in. Well, I want to throw in that happened on the matter most was this big really little explosion of fun on the fellowship of the link channel. Yeah, yeah. Which I think will end up possibly dealing with some of the goals here on the relate work. Yes, it's pushing at this. Anyway, so that'll just you know, but it was a lot of for me a lot of fun there. And so we were live on a jitzy call doing that that's why there was a torrent we were like okay everybody let's chat over there not in the chat over here. So I'll, and I'll put in the header of that channel, the time and date and jitzy link so that you all can join next week where we have a weekly standing call on Thursdays after GM, all that kind of stuff. And that that started because I was talking with Sam Klein SJ Klein, who's part of the underlay project with Danny Hillis. And, and he was like, we should just sort of have a general conversation let's ignore underlay for a little bit and let's go have this general conversation about this comments we're trying to build together of shared ideas. And by the way, I'm a little student today so all the things I'm saying are going to be a little off target but me, but close by I think. And then I met Flansian, who is like a mystery participant and has a funny ID and all turns out he's from Argentina and we spoke Spanish together for like the half hour the first time we met. And he's awesome and he's built this agora, which is very much in the spirit of Rome and Athens and log seek and everything else, but has its own sort of capacities and superpowers, but he's built the whole platform out and it's just kind of. It's, it's agora all the way down in ways that I think all of us would find pleasing, and then the tie that binds in the fellowship of the link. And which is a phrase I just came up with I'm like, you know, we're all into openness, and we understand the power of links and all that so that those are that's kind of the filter for fellowship of the link is like really working openly to build open platforms of some sort. And then let me zoom back out and find the edges a little bit. You're both very familiar with OGM and all of that, and OGM has become sort of a series of conversations within this large frame of how do we build both this infrastructure for sharing what we know and also how do we build the piece of OGM that I feel like we don't spend a lot of time on, and we kind of neglect, but not because it's not important is the softer side that has nothing to do with technology and everything to do with. Gosh, if I don't trust you there's no reason I should listen to your sensible arguments, or pay attention to data or anything like that. So that's that's kind of the box for OGM. And then I ended up in a conversation with Ev Williams, the founder of blogger, then Twitter, then medium. And I knew Ev before the invented blogger, because I was a consultant, I was kind of like an external advisor not not so much consultant just to an advisor to Pyra, which was the name of the little company before they before the guy across the room who was ever busy talking about things that would be familiar kind of in the wiki and collaboration world guy across the room and then blogger that eats the company over the course of five or six or seven years, which includes a dip where at one point, the whole company is one server in Ev's living room and everybody's had to be they've had to fire everybody because they had no business model. A year later it's three guys in an office on Market Street. A year later they were bought by Google and he's now a Google employee and all of that kind of stuff so so super interesting lots of history there. And I had a similar conversation with Ev when he was early on medium, and I was already like way deep in the brain and I was like, There's a thing here where we could build context for the stories that people are posting in medium wouldn't that be interesting. I mean, and it didn't sort of stick so much then but then we, I contacted him again, and we talked a couple months ago, and he was like, and he is now a fan of using mem. This is another Rome like tool that lets you share notes, john borthwick from beta works who's also a player in this puzzle as is because become a Rome fan, and is busy trying to figure out how do we do better than Rome. So, I'm helping him create a tools for thinking camp around this body of tools. So with Ev where the conversation went and this is, this is the midterm goals, part of the relate wiki is exactly meant to be this. The conversation went has has the simple goal of why can't we share notes which falls into the short term goals of like how do we prototype that how do we do that. So what's the goal of what would it take, and here, broad imagination, don't assume anything, but what would it take to, to build out. I don't even want to call it an infrastructure but it would probably have to do with protocols and API is and it would probably have to do with prototypes and examples, and sample code. It might have to do with legislation I don't think so there's no there's no thought of that right now but it might. And create a movement and some platforms that would that would build out a shared memory for humans without creating a whole new Facebook or a whole new Lincoln or a whole new one platform to rule them all. So preserving highly distributed data in individual privacy. And in particular, and I'm here I'm layering on some of my particulars about this in particular allowing different people to participate with their favorite tool, but still getting the benefit of the shared work in the middle. And also allowing people to express their point of view and their biases and their belief system. The Wikipedia for example doesn't let us do because the Wikipedia is as an encyclopedia is meant to be neutral point of view. In fact they invented this notion of neutral point of view and have you know, a whole bunch of culture, culture now built around that. And I'm, I'm a believer that we need to acknowledge and express our points of view as best we can. And if we do that faithfully and in sort of the spirit of discourse, but then we can compare notes. And we can end up complexifying issues and taking them apart and saying, Oh, we agree on these six things we disagree on these six. Let's figure out experiments or queries or let's bring experts in or let's do something else to address address those things. However, our assumptions are simply fundamentally different. Yeah, you know, this person's coming from a faith tradition where these are the these are the basic understandings, which just don't drive in at all with with the other. And then maybe the question is how do we find other ways to live together on the same rock without destroying the rock, which I think is an equally important conversation these days, and I've gone a little far field from the core of this. So the medium, the midterm goal part of this is kind of my charter with Ev on a short term project to figure out what norms organizations protocols apis and dynamics. I'm making up this list would get us into would bootstrap us into a reliable shared memory of the kind I just described. Did that make any sense. Yes. Let me see if I can wrangle it a little tighter. So, I heard shared memory. I don't know it's been a goal of you. The brain is one type of that. I also hear you talking about people's beliefs and how we can work on that and stuff like that. And for me, something I'm building right now is an epistemic map which is kind of a name for what Billy bot is, which, you know triggers that. But also, we also talk about note taking. So then I'm thinking. Okay, what's the what are the different types of data that we're going to want to have in the shared memory is it literally everything. So word Excel, notepad. You know, medical study data is it do we have to support any data format, and any set of structured versus unstructured on the scale, or is there a limit they're saying we just want to be epistemic sorry, or we just want to be linked data with text. Do you have a feel for the scale. Yeah, it's a it's a great question and it gets, it gets the warning quickly, sort of. So I think a piece of this depends on how much needs to be in the thinking system environment lens. And so there was there was a question as part of trying to come up with a think camp we had a little exercise to figure out what do we mean by tools for thinking what are the boundaries of tools for thinking. And, and I saw that that was very helpful. Yeah, and as far as I got with that I was like, Hey, people think when they write essays and you could use any simple word processor to write an essay I don't consider word processors thinking tools. So for me, you've got to bump up with a couple features to get into thinking tools. But if you took emacs I mean, not log seek but there's an emacs extension that that basically does Rome, it emulates Rome. It's called. Oh, I'll find it. And it's like, wow, somebody took emacs which is sort of self programmable and extensible, and they made it do what role is doing, we are clearly in thinking tools territory now. And so I don't know what the minimum list of those added features is but it has something to do with Wiki links everywhere or, or backlinks that has something to do with tagging and metadata, it has something to do with visualization and mapping. It has something to do with a couple other things like that. And some tools are thinking do zero mapping but they do these other things really well. Some of them, like the brain has no machine intelligence in it whatsoever. But it's really good at the mapping thing. Right. And it's pretty good at the labeling and tagging thing man man man. And so one of the things I came up with, which I'm not sure I have handy. I did a spider diagram I think it's in one of the pages I put on the on the wiki. I did a little spider chart or radar chart with dimensions that are kind of the things I'm naming right now. And I thought of that as a way of rating and then picturing different kinds of tools that one would consider tools for thinking. So, so sorry, long digression about what an earth is a tool for thinking in order then to come back to what do we want to see when we're thinking together. And I'm telling that getting down to the raw data is really pretty important that if, if the tool for thinking were meant to be only the abstract thoughts, if, if it was meant to be like, did Cartian logic or, or Hegelian logic with only the logical statements in a perfect crisp map, without any evidence without any, I wouldn't like that I wouldn't use that it wouldn't make sense to me. One of the things I love about the brain is that it lets me mix raw data like articles. Things with insights derived from articles, which then point to a whole bunch of different articles and evidence. So this argues for, hey, at some corner here we're going to want to have an R database running in in Jupiter notebooks or whatever and we're going to want to have access to raw data and understand which data is reliable which isn't. And that's going to take some expertise in data management cleaning, etc, etc. And I'm not going to do that today, but I'm certainly that's certainly on the table for what could be included maybe ought to be included here. And, and, you know, wiki data and a bunch of other open databases are all over, figuring out how to do collective intelligence collective sense making around the actual data in our world. And I think that's brilliant. And I want to include those communities in this inquiry, because as they come up with experiments that they can then test and do analysis on, and they can then come come back with conclusions. Those could close the loop and answer some of the experiments being set up from a more philosophical point of view, for example. So, does that make any sense. Yeah. So I'm not foreclosing a lot of data types I think I think I'm saying that he said, I think we're boiling the ocean, which is sort of yeah yeah that's okay you know you bought the ocean one bucket at a time right. So where do we start but I but I do hear some kind of principles. Like being able to bring any tool. So the having a principle of separating data from visualization, and having the data available so for example let's take massive wiki, and you're really wiki site is that if, if you posted the raw markdown files on the server, and there was some easy way for someone viewing the site to know how to do it, and then they could download them and put them in their own city and or go to get hub of course which they're there and pull them from that, which is very much the wiki, the massive wiki model. And we make that a standard because I was thinking my website is run from an air table base and some markdown files, which are hosted on GitHub. And on every page it should be, there should be a button to say, how do you get to the data on this page and maybe there's a standard like you do it.JSON at the end or I like that better than doing the having to set. I don't know, Jerry how much you do with the web stuff but you can, you can set application JSON on a web call but I can't do that in my browser. So I want to have a way I can do it in the browser. I want to, to have that kind of it as one of the basic standards that you always post the raw, the raw data is always accessible. And then we start saying here the suggested raw data formats. So there'd be we need to kind of figure out. And then we kind of need to figure out is it all going to be kind of static data, or is it going to be like here's API. So when you talk about Jerry's brain. Is there a format we could store it as raw files on the web server that someone could navigate, or do you have to have an API for something that big and not complex. And then are we going to support API's but I think the main thing is separation of data and, and editing and visualizing. So when I look at that data it'd be nice if it had some clues in it to tell me what formats this is, what's the suggested editor or editors what's suggested visualizer or visualize. Yeah, or even just what was this originated in so that you know what the optimal tool was. Yeah, exactly. And, and it's very easy for me to conceptualize separating data from tools from visualization or analytic tools that that's like, check, no problem. Once we get into protocols API's is the data live or not etc I start getting like a little, a little float in the ocean because it's more conceptual, and it feels to me like we've already exported my brain to a big bag of JSON objects in a postgres database. And if this were, if this were the base of bits that I was using I could keep it up to date all the time, and it would just live out there as a bunch of tuples and in some database and, and at some point it would be a distributed database and who knows what. But that makes sense to me, right. And then any node could be open for somebody else to contribute to in, as I envision it kind of a fork and pull kind of mode, which is one of the nice reasons to use GitHub under the hood. And that it bakes in, it bakes in a group collaboration, not necessarily a mass or large scale collaboration method but at least a community collaboration scheme that's proven to work pretty well. Sorry, Bill, I'm just getting a lot of these things out of our heads and into the conversation. Go ahead. Okay. So yes, I mean very, I like the conversation so just a little trick. Massive wiki. You're looking at a page. Just type md at the end of the URL and it cares your turn. And you'll see the markdown page. Right. That's just one of those just how it's working. It's just because the way Pete and I do this mess work you build the stores the markdown. In the same path in the same path. And yes, right at the same place. That's great. It's, I mean it doesn't answer the questions that Bentley wanted a little bit of medic information but here's some, you know, here are like some cordial things to know about what you're looking at, you know, to solve them to provide the following affordance. So, but nonetheless, that's just a mess of wiki thing. So one thing I want to throw out a little bit. I'm really kind of picky about language a little bit but I want to throw out when we talk about data. I think we should retire the term raw, because in the world of science there is no nothing is raw and if you're using instruments to collect data. There is nothing that is not processed. Literally, because the moment you collect some data you've the instrument already has has built like, you know, a digital. You know, a digital thermometer has already built into it a model of temperature. Because it's sampling. For example, doing computations on whatever signal is being. It's transforming. Anyway, so I used to use the one I was teaching on data informatics I use the word primary data. And NASA I can think this up and NASA has a really nice chart of levels of data processing levels zero through five. Anyway, where they really talk about, you know, data that's zero levels basically you know we got rid of all the all the noise, the obvious noise from the instrument. And then there's other process so they have levels of qualifying. The transformations have been done on the original signals that were collected. So it's nice use the word primary, then because people can talk about primary. One could make a definition for that. I mean I guess you could make a definition for raw but it's, I would like to see us just be more just establish a glossary that vocabulary that really helps us. I made a comment when I was looking at Matthew Lowry's big medium page. You know, so my pre association note was like, are we talking. Is it jargon versus versus the general tool, the general idea. I think we should. And I was just kind of flip but I really just had this thought we could do that here. I think as we go forward, we could just try and create a little definition of the terms are using. And, and also the one last thing I want to throw out I posted this in Manobos several months ago because I read this really interesting little note, but writing. There was an author who actually wrote a really nice essay that said writing is thinking. I could argue anything that allows you to like, you know, put inky scratches on a piece of paper is the thinking tool. Because until, and it even happens, you know, when you try and talk create the, you know, a simpler compound sentence about a thought in your mind you're actually thinking. Because before that it has it's not really even. It's like, I don't know, not expressed in any way so I don't want to push back I think you're right that we have these other tools that add more capacity but to say that if there are no links all you did was type, you know, use your typewriter in a blank piece of paper. You'd have to say yeah that's thinking creating the type of paper. I just created a glossary page and put in a summary of what we just talked about I think is this what you're thinking about. Well, I think we as we talk as we go through I think it would be helpful to always ensure that we are all referring to, at least a close enough concept or notion about what we're saying. Because it's very easy to diverge. I'm not. I'm okay with not using raw I don't know if I like primary because when you're talking about a shared data system, this primary mean Jerry's original page or my copy of Jerry's data. It sounds like it sounds like the original. But I mean if we just say data versus visualization maybe that's still when I was teaching informatics but I was teaching about scientific data so it was important to that was a useful descriptor primary data. So I'm not wedded to it, but I just think what I mean I think it's broad data is out there people going to use it we just have to be know know that it's kind of a loose term. Yeah, well yeah I mean so I use raw data a lot and it's not for me it's never meant the original unprocessed data it's just before it's been manipulated into visualization and you had you had some loss. So, you know, an air table basis still the raw data because no data loss, but when you put it on a webpage there's data loss. So I mean I'm not going to push this because I'm talking from a scientific mean you know when you're talking when it talks about NASA about astronomical data there. They got a list. It's one of these five pick one, you know. And so so Bentley data that were the outliers have been thrown out where the decimals have been rounded where I'm just making things up but but five things have happened to it. But it's but you still haven't done anything to analyze or visualize it that still raw data to you. No, I wouldn't want to normally take the decimals points out and stuff like that. Right. Yeah, I mean that was a bad example. But yeah, yeah, I mean, if you take. Yeah. Yeah, I don't know it is it is Lucy there is no you know the human language doesn't isn't specific enough, but there is a look you can take if you took raw data and then you put it into a you know, a data format. And I knew that raw data that that if if you took, you know, a scan of someone's handwritten notes you took that you try to put that into a data table based on the scan. And yeah, technically the raw data would be the, the bits before the scan, but there's a fairly do we go that far back. There's a tiny interesting story here about a you see at my alma mater as you see Irvine, and there was a professor there named Sherwood Roland, who discovered the hole in the ozone layer. And the way he did that was he stopped throwing out the outliers that other scientists were like now this can't be. And the outliers were in fact describing or a tear in the ozone layer. Yeah, maybe, maybe source data is like what's the source data for this visualization for this visual. That's good, that's good because then someone could go if they found it. They could ask questions about the quality of that information. Although then is it the original source or the copies. In the digital world there are no originals I mean literally if you're doing bit for big copies from just this service. Yeah, it's really is not original. So it's really the source for this visualization or what's your source data used that you manipulated to get into this age age. So, so this is just going to throw it in from a little time with the dynamics heat thing, you know, the deal with like decimal points. The reason you want to round them off is like, you know, people at digital thermometers, somebody, you know, somebody gives you a temperature to four decimal points that's like, I don't know what that is like. Yeah, it's not that is not a right. You know, it's not a value. So we've gone pretty deep down a vocabulary and terminology rabbit hole. I would love to bounce back up to the picture of the midterm goals and how to describe this thing in a way that's kind of pity and intriguing and invites exploration and helps frame that this kind of heart of the relate project in some sense so all thoughts welcome and I'm going to screen share and go back to that page in matter most and just take not matter most in I'm sitting and just take notes there. Yeah, so have you looked at Matthew Lowry's little medium thing that he's writing this like he's got this big book going on in his personal knowledge graph. I skim the chapter I haven't. I think his picture really made me think because it's very generative. I looked at the picture and I didn't get the same generativity you got this is the picture. I did I had a thousand questions. You know, but the, but the thing that he says that's interesting is he has some, I took some preliminary notes. And I'll probably look at some more but he touches on me. It's obvious he's in the same space he touches on some of the same issues about trying to share, create something that shareable, being able to so that there's a lot of similar terminology here. I built this. It's a framework model for. I mean the one thing I liked about Pete Kaminsky is using the word massive. When it was, you know, I'm a SVF was that you know when this with the world be like we just have this just a few basic things. We're using markdowns. Things are shared. The version control. And because you're using computers, the files. That's it. Right. And the question of, well, how do you share them anyway? It's like, we're talking about, you know, these simple formats. So I found that to be makes that makes the massive wicked thing really easy. You could put it almost anywhere because basically just a bunch of asking text. So I just think I find there's some similarities here and this fits into the, you know, it's what Yuri has talked about. There's, so this is a very many people are having the same conversation. And in, and here I have got a page called relate pioneers. So I need to call this Excalibrain, because I think that's what Pete wrote about in today's Plex, which makes more sense. And each of these needs to be a link. So these are a whole bunch of projects like this that I know about and I should add to this list. So. This is a Gora, the show is doing Excalibrain and some other stuff. This is rich burden who hasn't really, he's hard to get a hold of. But, and he was the CTO at the brain for one year along like 20 years ago. And then he went off and did a bunch of other stuff, including a company that Pete was working with wire line I think it was called. Anyway, he's created this project called the distributed operating system which works like I did a demo for me and I think a year ago and then it was really pretty impressive and pretty it was sort of pretty like Google draw like Google Docs is pretty without this without all the features of Google Docs but it was but it was an elegant environment. I don't know that obsidian I would call an elegant environment that was better than than emacs for me anyway. This is Pete's massive human intelligence project. Matthew Lowry social knowledge graphs vision, which is exactly the page we were looking at curies. I don't think trail marks is the name he's talking now it's hyper hypergraph. I'm forgetting exactly what the latest rev of the curies projects name is, but this is this is a batch of them and I think I'm missing a couple that we know of in our communities. But for me, I'm trying to figure out for the middle for the for the short term goals of relate. I'm trying to figure out how to throw these people into a comfortable salon together to remix their ideas to start by just sharing notes across the different tools as much as possible. Generally, if possible, but pair wise if need be, and that's a good that's a to me that's just like an interesting starting point, but then to compare notes and say and compare architectures and say wow we seem to have built the same thing only slightly different ways and make a couple choices and maybe unify some projects, I don't know. And out of that to start thinking about these larger framing issues about gosh, we could do this except there's no API or protocol for X to do to do some things we someone somewhere needs to stand that up. And that would be great because then we can say oh, we can create an engineering effort to stand up the protocol and maybe an organization to host it, or find an organization that already has similar protocols. I'm kind of wandering in that space. And I know that we've got a bunch of people on deck like in the conversation already who've gone and build some some big hunk of this thing that we're aiming for that makes sense. So, so the midterm goals. The top part here was basically the skinniest version of this I could think of so what will it take to create a large scale memory that is crowdsourced like Wikipedia, but doesn't necessarily have Wikipedia's model for existence could be any model that allows persistence use multiple memory and context making tools that allows participants to express their own points of view that encourages the crystallizing of those points of view. Not seven billion humans each with their own completely different than separate point of view but rather. Hey what what what Matthew wrote about this part of the business model totally speaks for me. I'm going to point to that and so will 60,000 other people. That's what I mean by crystallizing. So here's a question for me. So crystallizing. I look at. I know this probably puts it on a weird, I'm going to live as an engineer so crystallizing as a solution. What problem is this solving. I know what what Bentley's been working on and trying to help people try to organize argument and put together, you know, whatever the kinds of ideas evidence and stuff that solves a problem that many of us have right right. It's not going to be easy because you know, I already have an opinion don't confuse me with the facts. But this. I look at this is the thing would be for the relate project to say, here is how we're going to. Here's how we're going to relate our learnings discoveries hypotheses, prognostications we're going to use the following tools with the following interconnecting, you know, processes practices. Because I in, in a way this list of all the stuff that's happening is like, someday there's going to be a small number of that really seem to work for which everybody's like okay, give it up I'll write the software like this. Who cares, right, it'll work. I can debug it when it breaks. I'm happy. So, I, I just don't want to mean it's like, when two people have a problem they're more likely to try and find a way to solve it. Right. And if three that's maybe three people like well let's you know instead of doing it by hand, let's make that easier. So here we've solved, you know, the problem that people have when they get together and can't figure out how to agree it's like, you don't know about that one. Yeah, I'm trying to make a dent in this not necessarily solve it in some canonical fashion and some central fashion I think I'm trying to figure out, how do we work toward one another. Okay, so what kind of a, what kind of a knowledge, what kind of a knowledge collection. What's the artifacts are going to be left after relate gets through its medium goals, will there be an artifact that actually can be searched from which things can be. Yeah, yeah. So, so, so the nearest experiment like this that I can see happening quickly is flance and Zagora creates markdown pages on GitHub. And so does the wiki I'm working on and if we export some chunk of my brain out into the same space, we can then start to overlap with each other around the same sorts of nodes and thoughts. That to me is a, is a like a very, very simple use case that's interesting. And from that we can start to get to some of these things. Okay, so I think so that would be a short term staying one of those things up. So what I mean in the short term goals here about taking the relate pioneers and build some prototypes so that that was where I was heading here. So it's sort of what I feel like what's happening with the massive wiki kind of thing. We have some things organized we have some tools that are now, you know, and, and just to keep going to work. So I'm going to summarize on the big picture also. And I need to put in here. A new turn a new item here. Just to keep to keep the big ideas also in view. I think that we're heading helps change how education and journalism and science and governance all work. I'm very disconnected from each other right now but what if educational projects for feeding a shared memory, like Wikipedia but different that journalists were also feeding and feeding off of. And when scientists did research they were putting their work someplace where it was inspectable replicable reusable by other people, and then that turned into books and papers and whatever else. And by rethink media I mean the video I shot about hey PowerPoint is a playlist books or playlists. Let's rethink the constituent parts of all these ideas and how we share them, so that we can bump out of emulating magazines and movies, which is what we're doing right now. So, if we could jump back to the midterm goals. Maybe it'd be interesting to step through one of the examples like that plantia. Because it's the it's like the third and the fourth bullet point that I think we start losing traction maybe. And midterm goals so that these may be more long midterm goals and the first couple maybe the simple thing to do. So we still get into, or I guess, there's some, there's a nuance there that I'm wondering so in theory, if I have, you know, let's let's say, plants, you said plants is data was on GitHub. Yes. Okay, and then let's say, let's say we exported part of your brain to get hub. Yes. So you can go and copy a page from plantia Zagora and put it in your, it, you know, copy it into your GitHub. You could also put a link to that. Right. But then there's no way for people to see both, or to know that they're connected. Well, that's metadata right. That's the thing you were saying earlier about hey where did this where did this data actually originate from who created it using what tool that needs to live as context for the data somehow and there's Nora Bateson's whole warm data contextual data thing. I think those are important factors to include here. Well, it's even it's a little harder than that because if I don't, if I don't know to look at your. If I, if I don't know to look at your GitHub I won't see the metadata for your GitHub to know that you have a copy of flancia unless flancia has a has a pointer. Unless we, unless we decide, we do the best way to do this is to head back for toward canonical representations of the data. And then each of us is layering on whatever metadata or extra stuff we need to make it work. But until and unless that data changes fundamentally, we're always pointing at the same ball of twine. So even if you don't do canonical, if, as long as you have a, a directory and you could have multiple directory services so it's kind of Google for this thing so it's, it's a site that either crawl is crawling both your data and reads the metadata and says oh these are these were the same item and here's a path, or, you know, as you copy out some of the data for your own version you be reported to that. And then you can centralize but multiple centralized search. Yeah, tools. Agreed. Yeah, so that would be the crowd source in the Wikipedia allows participants to use more. So in since we're working on the raw data and not on the visualized. Since we're working on the data and not visualization. And we're kind of, we're hitting that although it'd be nice to have the metadata to say what the tool was a built in. Yeah, or what and and slash or what data, because some of the tools have multiple data formats so. But yeah, some way to say, what's the shape of this data. So allows persist participants express their own point of view if you have a copy of it, and you've edited it. Is that expressing your own point of view. Is that sufficient or is there more needed there. Sorry, so again. So, now the fourth bullet allows participants to express their own point of view. If you make a copy of fancy is one of fancy as items and then you change it that is you expressing your own point of view. Yes. So kind of have that. And then, and that's, and that's totally complimentary to this crystallizing thing because I might agree with once again about 80% of his policy position on HTML. Whoa, is what just happened. I think it looks like their browser just kind of rebooted all the pages. Hold on. My browser just rebooted itself, and everything started playing. What's there we go. Wow, this is fascinating. Theater of the absurd. My browser just rebooted every tab. Wow, okay I exploded something here. Hey, I'm back. I have a question I have to jump off in a few minutes. I wonder, do you all, we need this work really needs connections with the basically the library and information science community. And the citizen science community, especially moving forward to the long to the long term, because there's a lot. A lot of established working knowledge and practice there. Yes, and we're actually connected into some of these communities I mean, you know we know people at the archive we know people in different communities doing stuff. Especially with the data stuff because there's a lot going on in the International Scientific, the International Council of Science on managing scientific data I spent 10 years with a project working on preservation of scientific data so there's, you know there are a lot of efforts and there's a lot like the AGU is a geophysical union, you know the one that that it's got a Virginia members. They have been doing an enormous amount of work on data because they collected everywhere from the ocean and the air and the land and a friend of mine from Woods Hole Oceanographic was their data collection specialist from mostly underwater sensors and shipboard sensors and all that kind of stuff really cool. Anyway, so there are projects it would be nice somehow it's once this if you get some legs under this to see what's happening there because sometimes I get the feeling that some of these technology driven. You know I'm going to start with semantic web and you know hyperlinks it's like you should go take a take a walk and talk to a few people. You know, because Well, I think, I think that a lot of this work is being done in neighboring communities, where some of it's just really deep work that that they we need their experience and wisdom. And if we can, and if we can figure out what the loose scaffolding in the middle looks like, then it's easy. Well, relatively easy to go say hey. This is how that this is how all the wisdom coming out of data collection and science fits into this model for information sharing. And then somebody who knows what they're up to goes and figures out the details of how those things marry up. But I think I think what I'm trying to figure out is how what what is this bare bones skeletal scaffolding look like it's a little bit like this is a terrible analogy. I know how to make internal combustion automobiles, and they kind of know where the transmission goes and how an axle looks, and where the differential goes and whether there is a differential and sometimes the engine is back here and sometimes it's up here. Now we're in the state where we're going electric and for the last decade or two there've been these deep conversations about, do the motors go inside each wheel. Do we have four motors on every card of the other four tires is the chassis flat or shaped. It depends on better, like, all of the questions about all of our old assumptions about how to do stuff around a car and mobility are being rethought and in some cases improved on I think a lot. And that's going to change everything we thought went on top of the wheels of a car, right. And for me it's just going to be pods and now we're going to be able to design pods like will there be a massage pod there'll be a vaping pod there'll be a brainstorming pod there'll be an isolation soundproof pod so that you can have quiet time. And that's really cool. And I think what we're doing here is sort of metaphorically similar in the sense of, man we're trying to sort of rethink, think our way away from internal combustion websites and books and magazines into this new ish space of collaboration. I don't know why I'm here because Doug Engelbart showed it to us in 1968 we still don't have that. But, but how do we bring us about using more or less modern technology. I don't know if that whole riff was was was useful, but for me it was like yeah yeah we're at this moment of possibility and reinvention. And a lot of communities around the world are taking bites out of the different problems involved here. And they're not, and they're not bridged or talking to each other very much. And I think a piece of the longer term relate mission is to create a happy harbor or happy docking dry dock where these pieces can come together. I've been watching too much sci-fi I thought you're going to say docking station so sorry. I got it. I got it right. It's almost dinner time Casa Anderson here so thank you very much this super helpful. Yeah, no thank you all and it's great to see you badly. Good see you again. Happy to stop happy to keep going I don't know what where your head is on this. I want to finish that thought so I think one of the infrastructure technologies. I mean it is kind of like back links, it's just back links is a specific implementation. I'd rather go another way to where there's this sort of infrastructure of a pointer to all these different versions of the same data, or people have their own expressed versions. And then later on we can work on what's the process for crystallizing these. That's a different ways you can do that, but the question we need to answer is how do we want to know when someone makes a copy and make their own viewpoint. How do we know that those two are related and it's slightly more than, yeah the med just having metadata on it when you make copies useful and then telling the person you made a copy so they could put a link to it if they have time. I think it'd be better to have something a little bit more, a less, less manually intervene so having something like a Google that is running over all this data and creating a directory of all these. Is content addressability solved this kind of. Well, so, so maybe, I mean, well, sort of. So it'd be nice if. Well, no, because the thing you're doing is a different chunk of data now because you've put your own view on it. It's not just a different view is you're expressing a different opinion with the data not just about the data. But a different version of that data so the data is different so the hashes aren't the same. So somewhere we still need to record that this day is like this data, and it needs to go both ways, not needs to be nice. So V one could be if I make a copy the standard is in my metadata I say I point to the source. But I think to get to that foot fifth or six bullet point you'll need something that aggregates all that metadata and to a few central places that people could search. And, and we're, we're getting into issues like transclusion, like what is proper transclusion, how does that work, but also there are some hacks hackable solutions in the interim like fed wiki basically super replicates everything. So, if I go to your fed wiki and and go start reading a page, my understanding of how it works is it just copies that page over into my fed wiki. Yeah, and says and says thank you very much. And then I'm working with a local copy whatever I do with it. And then I can, if I were to make a change to it I have no idea whether it notifies you and says would you like I have no idea if there's any kind of fork and pull there. To like replicate. What's the word. promiscuously. Yeah. Right. Yeah. And it's kind of a short term way of addressing this issue. I still have the problem that's where it's really hard to see the complete picture of all the people that have their own on this thing, which could be a later thing but the having built building into the standard that you always, when you make a copy always say what the sources which may be what they do and fed wiki. And then, you know, there's a standard thing to communicate back to them to say what it is and then there's this, you know, and then we could build on top of it later on at some point the technology to give you a view of all the different versions. Yeah, so maybe it's just a group of standards that says, you know, I created, here's the data, not the view. Here's the tool I'm using to edit it. Here's the data type within that tool. And you can figure out other tools that could use that data type. And then here's the source of the data, the pointer to the URL that got the data from. And then you kind of have the raw workings of kind of minimum viable product. And in GitHub, I don't understand yet the difference between forking branching and cloning. And I think branching is considerably different from forking. Yeah, so well. And I'm not asking that because I want to understand this is right now, I'm only saying, these are different ways of taking a copy and doing something about it and notifying the originator that that's something interesting might have happened to a piece of something that was created. Yeah. Yeah, forking branching and cloning are all taking a copy. It depends on how far they are away. And whether you have permission to merge in. So you're either then merging or doing a poor request, depending on your permissions. So I'd be happy to go over that with you sometime that's not that hard once you get here. Now, for the most primitive versions of what we're talking about. I don't know that we need to solve this issue of the data synchronization or doing like data synchronization. Well, I think data coherence I don't know what to call it exactly. Yeah, or are you talking about the crystallization part or just no, the crystallization, the crystallization I think I, I think it's a bugaboo in my head I don't know that anybody else cares. And the only the reasons I have for crystallization are a to prevent fragmentation and just replication all over the place I mean, well, actually, one of the goals of the system is to try to not minimize but not and not optimize I don't know what to call it replication of effort. So, so, so how do we prevent lots and lots and lots of copies of exactly the same thing happening all over the place, or even were slightly different things that we don't intended to be different. Exactly. And long ago there was a website I forgot which one, which had a couple of simple things, which took a folksonomy of tags and tried and were quite functional at narrowing the name space for that folks on me. So it was a plural and a singular it would identify that and connect them, you know, back together. A bunch of things like that which made the folks on me much more useful because it wasn't all broken and fragmented it tended back toward some some terms that started getting a lot of energy because one of the cool things about crystallizing or collapsing back in is then you can say, and this note has 66,000 people who have attested that they love this note and are pointing to it as part of their, you know, logic mainstream, that's a good thing. Right. And for me to play this out is, you know, David Reed would go to Congress and say, see this note right here with this behind this policy issue and this policy page. There's 2,336,224 people right this moment, who will vote for it. If you will approve it or put added to the bill you're you're sending in, right. Yeah. And I think that's really powerful as opposed to some lobbyists saying I speak for, you know, lots of people. I think the minimum viable product is a metadata protocol and some patterns. So, like, you post your data at a URL on the internet. And with that. I think that's your original information. Someone else wants to express their own view is that they copy that data posted at their own URL, and then they. Oh, I'm sorry. So in the first one you post your data, you post the app you used and you boast what type of data from that app it is. Someone can copy that, and then they are supposed to have the those the leave that metadata but add to it the source. And maybe the date and time it was taken. And then they can go in and edit it and it would override saying oh now it's done this tool and it's this data format. And so that would that would be, I think that would be minimum. So like you could do this with wiki with massive wiki. If we just on all the markdown pages there was some metadata or another file next to it that says this is using you know for your relate, it would say this is a city and then it's a standard city and which could easily go in YAML in the header and then there's a couple of plugins like data view and a few others that pick up that data and can do stuff with it. It's interesting so so that would be minimum. And then after that you're talking about building services and infrastructure to make that more visible and easier to do. And then you'd have apps that are compliant so if I wanted to make a new version of blogger. I would have it be a web editor that anyone go on and do it but in the end. Anyone can can put md md at the end of it they'd see the markdown they'd see the tool it was edited and blogger 2020. And then they could copy the data but the source back. And it would, you'd kind of have all these benefits. Makes sense. One quick thing just to kind of completely derail you from that thought on the relate wiki the tiles. One of the things about having like a little directory of things is that often the name for me isn't enough to be useful and then I click through to see what they are. Right. So a single sentence explanation of kind of what it is would be helpful and but that wouldn't look really good probably in the table format you have. Right. I was trying to figure out what I would suggest and I think one of the things I would suggest would be a. I don't know for this. Okay, would be. I don't know if you'd like this or not but have one row with all the columns in it and then a bottom merged cell that has like a description and then a spacer so then it's more like cards. No, it's just bros. Now, I don't know how the TV tables work very well but I'm a, I'm hoping that there's some reasonable word wrap and stuff like that so they're, the primitive answer is like just do a 10 word description of each project that fits in a cell and in the table. Yeah, I mean you're already getting bad rapping on that right. Yeah, shed. Yeah, that's only visible on this page I think it I think it's prettier when it prints to a web page. And until you view it on your phone. And then you get the exact same thing because they I think they both use the same. Gotcha. The same renderer ideas date tables are just hard on small screens. Yeah, okay. So, so there's probably a need for a different. So what would be need is to have, I don't know how useful this would be but it'd be need if that was data, and then when it was on the wiki site, it would say, Oh, this is a table, let's make these, you know, let's add a little JavaScript to make the column sort of a card view and, you know, instead of the table view, and give features like that, that that would work with any table. That's, yeah. And I'm correct me if I'm saying something that's not similar to what you just said but there could be metadata in the mean brain brain page the brain brain phase page etc etc that has a full text description. And then this table could just be a little app that pull that says hey, go pull these fields from each of, from any, go find any markdown page on this directory that has metadata that says it's a tile, and then sort them by this field and then pull these fields in and display them as cards. Yeah, I mean, even if the standard for wiki was that if you hovered over a link, it would, it would give you the, the text, longer text, the old text link, and then if you hovered over the mean brain one then you'd see what it, what it is. Gotcha. So, so, there's a lot of ways to do it. All of this partly to say that you're not a fan of clicking through to read what the thing is you would like it to be summarized up here. Yeah, I guess what I'm saying is yeah this, well this table is completely useless to me. Without that one little piece of information. If I don't know what mean brain is the name doesn't mean anything to me. I have to click through each one and I'm like way too lazy to do that. So I was thinking for like bringing McBrain face, which is an exact, what's that really is meant to be as an example open source visualizer of mean brain so it's not meant to be useful it's just a demo. Demo visualizer of mean brain would probably be the smallest like those four words that that would be useful, but no oh that's a demo I can go click on it learn if I know the code then I can adapt it or template. I don't know which column you'd want to put that in. Yeah. And there's also. And I don't know if you could put icons on there for the status instead of the words. I guess you could do either ASCII or emojis. So for which thing for like the finished and the proposed if you want to make a make a check mark for the finished and light bulb for the proposed. Good thought. Let me see if I can. So I just put in a new column. Yeah, description or something. Yeah. Yeah. And then meme brain is a Jerry's brain crawler. What is it. I guess I'd say an API for Jerry's brain. Let's see if that even let's see if that's even pretty. That's not terrible. Okay. Yeah, that's just not terrible. Okay. And we can replace the these guys the finished and proposed with icons that'd be good. Yeah, this gets ugly really fast. Yeah, tables are not. Yeah, not pretty. So much easier in a spreadsheet. I mean, you know, this could be a spreadsheet or an air table embedded and I get in a, in a full with the way editor in this. Yeah, I wanted to know for time so it'll just be like you're editing it. Yeah. But thank you that's helpful. Cool. All right well it was interesting talk. Thank you very much. Very fun. We'll keep exploring and see where we can get it. You were when you were proposing the problem of, you know, not being able to kind of like make decisions together and having all this high polarity it's like that's exactly what I'm working on. Yeah, you're at the kind of higher meta level. I'm a specific use case. And what I'm trying to figure out and this is why my brain hurts is what is the largest framing for a general purpose scaffolding that lets gully bot and other things you're working on fit elegantly next to everything else that's going on. Yeah, yeah so finding a way that I could post the gully bot data in a data format. And then put those pointers on it that'd be an interesting little experiment but you got to figure out the granularity level because it's like, I have a debate. Do I want to post the individual claims separately, or the whole debate and then how do they know. So yeah, gotta figure it out. Yes, it's confusing but I'll be thinking about that on on my project and sweet. I think of anything sweet. I'll see you in the next time.