 Recording is on. And the recording was very weird and janky. It came out kind of echoey and stuff. It was a little strange, so. But there was a recording last time from this. Pete, we're trying to troubleshoot wide. And Jitzi just changed how it does recordings. So that's different. So there we are. Hey, Mark Antoine. Hey, Mark Antoine. You've made it. Finally, I've managed to assist my other meeting. Cool, thank you. Flansen, do you have audio back? Can't tell if you can't hear it. It's still working. OK. That's strange. Is it not working? Yes, no? Huh. I think it never worked for him. Oh, OK. Jack, nice to see you. So I was just reporting in that I had a very nice kind of first call with Ida Josefina this morning of the SANE project. And we sort of told our stories about how we got to where we are. And she's trying to create collective intelligence so that extinction events don't like extinguish humanity. And I'm like, oh, that's me, too. So very nice. And I passed her a link to this call if she has time to join. I don't know if she has. She's in the Betaworks cohort at the ThinkCamp right now in New York. And that that's apparently going really well as well. Yeah, so this extinction thing. And she mentioned ex-risks, which is not a book I'd sort of found yet. And so very interesting stuff around that. And I now have, officially, way too many tabs open to harvest after all of my calls are done today. But I just want to see how everybody else is and what you'd like to talk about in this call, since our crowd is kind of mixing and mingling. Chris and Valencian, Mark Antoine, and Pete and Jack are regulars for the Free Juries Brain Call on Mondays. And so I think these two groups are very overlapping in terms of intention and context and all that. So happen to see us cross-pollinate. And we'd love to know kind of what's on our minds. I don't know if you saw the link I put on a SuperMind design. Just about a minute before I logged into this call. OK. Did I also put the YouTube link? Yeah, I did as a reply to it. Sorry. What's a SuperMind? Blah, blah, blah. This is somebody who comes from the MIT collective intelligence lab. And he has a site which he says is about augmented collective intelligence. And he's interested in how do we have SuperMinds. And interestingly, if you follow the SuperMind design link, you'll find somewhere. This is Gianni Giacomelli. Yes. Yes, yes. I've got him in my brain. So keep going. OK. There's a slash database gives you. He's got 800 examples of real-life building blocks of collective intelligence. Let me put that in this conversation. And he's got four line axes of how he classifies collective intelligence building blocks, I guess. And so he's got his own classification scheme, which I think is part of what you were trying to do in this group, and pretty rich typology and examples. So when I keep saying, God, we collective intelligence types need to assemble better. Perhaps a kind of what would you call it, collective intelligence, perhaps? Yeah. It's irony is a bound. So yeah, it's happening there, too. So that's what's on my mind these days. We have all these collective intelligence groups reinventing the wheel, and we need to talk more. I love that. Afflantzik, can you hear us now? Yes, good. Cool. I can, I think. Yes, can you hear me? Do we hear you five ways? Five by five, as they say in the radio business. I have it. Let me hear. And five by five means, Pete, you'll know this, right? I think it was five on five if you got five words out of five. I thought it was like a rating of one to five on quality and distortion. No, it's something like that. I knew it as a number of words. Five by five, I hear you loud and clear. Signal strength and readability report. Signal strength and clarity. Yeah, signal strength and signal clarity. Exactly. So five by five means the signal is very strong and the transmission is very clear. So there we are. As well. Kind of like an Aptgar score. Yeah, yeah, an Aptgar score for radio transmissions. Is it pink? That five by five is probably older. Yeah, yeah. I, Mark Antoine, you look so familiar and your name is familiar and I'm guessing maybe Tinderbox. That could well be. The pictures of all the post offices. OK. That could well be. All right. The first Tinderbox meeting. The very first. The very first. That's the only one I was at. Seriously. OK. Wow. What is Tinderbox I have to ask? It's a very, well, not that ancient, but 20 years. 25 years ago. Tool for thinking. 2002. 2002? OK, 20 years ago. It's an app. So it's not a web app. It's a Mac app. That's why it didn't get as popular as it could have been. But it's basically got a lot of what's making people excited about TAN out today, which is prototypes, a.k.a. super tags, and agents, a.k.a. live queries, and plus very nice visual organization. So yeah, curated and written. One of the big old personalized management systems. OGPKM, yeah. Yeah, I think it's, if I remember correctly, it's also got what I think is an interesting pricing model. It's high priced, kind of unapologetically. So it attracts new users slowly that retains users over time and has the development budget to be able to continue making it better and better. Based on meetups that I've seen in the last year or two, its user base is all academics over the age of 60 primarily. And part of that is because of when it started. But it hasn't spread because it never went into the web world. So who buys apps now? I know. And it's Mark Bernstein who I met once years ago back in sort of that era and haven't bumped into since then. He was very active in the hypertext groups. That's where he gets his name. I invited him to do a lecture at SRI once, and he came out. But yeah, at least historically, it was Mac only. It's not like it's still up on people's ears. Like Devon Think, which was Mac. Like Devon Think. Actually, go ahead, sorry. I had Tinderbox and Devon Think on one of those old white MacBooks. And the hard disk crashed. And it didn't just crash. I sent it to the guy that cleaned up the disks from the 9-11, the towers. And Sam Francisco got all the disks. He phoned me back, and he said, there is nothing I can do with this disk. It is gone. So I had to buy a new disk and start it over. And I didn't buy Devon Think again, because it's expensive. But Mark remembered me. So he gave me the upgrade price to the latest version, because now I was on a MacBook Pro. Tinderbox, by the way, was a rewrite, a simplified rewrite of something called Story Space, which way earlier, which was one of the pioneer hypertext writing tools and a lot of early literary experiments and hypertext used Story Space. And it was a more collaborative tool as well, very, very innovative tools that are lost to history with the web. And Story Space was inspired by Stretch Text, which we've mentioned several times in free-trade spin calls. You can get lost in the Skate site. There's just so much there. Yeah, I just posted a link to the 1998 essay he wrote. And Story Space is 1987 software. No, the history of hypertext is extremely fascinating. But anyway. So Chris, we're sort of talking through an archeological dig of these kinds of software, the way you dig through common place books. Oh, with men, but even most of the stuff I know about it is by having gone back and looked at who was influenced by what from older textual presentations. So that's all related. But anyway, I don't know if the agenda is still writing a compendium of tools. And the archeology is interesting. But I'm really interested in the discussion about what are the classification axes, which I know you're having here. So Matthew's not on this call, so we should probably hold off talking about that. I can represent that a little. Matthew and Bill Anderson and I have kind of taken that on as a project. Developing a set of dimensions, five, six, eight dimensions, along which you would profile different tools. And then creating a massive wiki, actually, with a person page. And then here's the tools I like or dislike. And that links off to tool pages with scores and a spider graph and things like that. So let me dig around and I'll find a link to kind of the massive wiki we're using as a project coordination thing. It's really rough and in rough shape. But that project is ongoing. And briefly, before you go on, Pete, that's what I was going to talk with you about for the Plex is that you and I could sort of compose a couple of sentences about that project to put in the Plex. That would be great. I want to interview you about the gym, actually. OK. So where we are with that is I've volunteered to write some of the Python stuff that would help us interlink person profiles and tool pages and things like that. You can use regular wiki links, but then we'll have something that collects the wiki links into here's a list of all of the tools that are mentioned and things like that. Matthew, between this week and next week, is working on the particular dimensions. And I think he may end up using massive wiki to talk to some of the massive wiki channel on MatterMust, but I'll also encourage him to cross post a fellowship with Link if he's not started there already. Thank you. That's great. And Mark Antoine, the whole idea of the dimensions, the vectors dimensions, whatever we want to call them, that we're going to rate or evaluate the different tools on is important. Is there a place where we can hold that conversation or we're collecting up that information? You've muted. What I understand is you're having a conversation on the FOTL massive wiki, if I'm not wrong. I did propose a Google doc where I had my own thinking about axes. And I don't know how much you refer to that. This is something I've done partly in collaboration with the CDL people who are also doing the same thing. What was the title of that doc? It was. I'll give you the doc reference. Thank you, because I think it's coming together in my head that I need to piece these things together still. Yes. CI features and goals. OK, I've got it in my brain. You got it. Good. I'll put the handle for others. Thank you. And the reason I, as I said, we have now a new dimensionality proposal from the supervised design folks in the database link that I put just above. So I think we need to do some hard thinking about we all have ideas about the important dimension. Now, my document is way too detailed. And I probably need to get a simplified version somewhere. But I do think it's a contribution, because I've been thinking about it for a while. And I'm willing to go through it, because it's probably not as clear as it should be. But I don't know when. I thought today would be the time where we'd discuss these dimensions. My guess is Matthew's going to be thinking about it further the next week. I have a write up. And that might be a good time for him to start syncing up with your thinking, Mark and Tom. And I'm posting a Google Doc link to the spreadsheet that Bentley helped create a filter for, which lets you see what a radar diagram might look like using some dimensions. And this has not been synchronized at all with your document, Mark and Tom. I'm aware. Yeah. And Matthew is aware that Bill Sice has stuff, Mark and Tom has stuff. And so it's not like he's ignoring that. But I think just to give him a week of kind of, let me kind of drop what I'm thinking and then start to co-lust it up together. Fair, fair. I needed to do it for myself for one thing. And I understand that he had to do the same. Yeah. What I hope is we'll take time to align. Cool. Other thoughts around this particular project? Shall we open up the dimensions problem? I think we don't need to open up the dimensions right now. And that whole massive wiki that I posted is, if you click all pages thing, it's all kind of project plan and a little bit of starting a little bit of prototyping. And it's nothing ready yet. So I don't know that it's worth walking through anything even. I guess on that project plan page, which I think is the one that I talked about, it's like first word in the build phase right now, just start figuring out what we're doing and starting to work. So then we'll start bringing it to larger F, L-shaped link and OGM folks, and then keep going from there. Concentrics and cool down. That's awesome. Thank you. I have a quick random thing. There's a tool called Scrantl, which is getting towards the end of their preview early adopter pricing promise. $60 for a year and the early adopter pricing stays for as long as you have the account. So it's worth looking at and seeing if you're interested. I think if you're not interested in money at front like that, you've got on a wait list or they'll have a free tier in six months or something like that. You're aware of another one in that space is Fermat? And that the tool for thoughts for rocks has an event tomorrow and they're going to Fermat and another one. I didn't realize there was a call tomorrow. I'm somehow off the notifications for that. Because you're muted. There you go. I don't know where my window went. It's the problem with Gentile. I mean, with Jitsie, it's just a tab somewhere. B, I've missed the tools for thought, rocks, thing too. Which I'll probably miss for other reasons, but it sounds interesting. They always record movies, so. Yeah. Thank you for that. Thank you for the mic. So since we're kind of on this topic, which of these rocks do you want to turn over, if any? So it seems like there's a lot of tools in this space now. Tana is somehow a neighbor. There's just a whole lot of visual whiteboards, a zoomable whiteboard, lots of notes on it, maybe links, maybe machine learning and other stuff. I don't know. But there's a proliferation of this idea all over the place. Go ahead, Funcion. Oh, so the unit is wrapped. Oh, I'm done. But yeah, I guess. So my particular goal interest is in interlinking and just networking tools. So I go back to the dimensions and I know this is going to be sort of like dimension charts. I'm personally interested in exploring how even just a subset of the tools that we know of already could be made to interrupt or to interlink. The sort of the hypothesis that interlinking even a subset could allow all those communities to compose and perhaps actually push them towards critical mass in a position perhaps to some less networking-minded first mover. Not to be offensive, but just to try this hypothesis. So I guess I'm personally less interested in having a completely thorough mapping of the space although I believe it's very important. And more of all, what could we do? Perhaps discussing what could we do with what we already know. Well, that's my hyper-knowledge plan is all about that. But it makes quite a few strong hypotheses that what we want, of course, we want deep links. So that means we both need a way to access objects and different tools for thoughts and to point at sub-objects sometimes maybe like phrase in a node, things like that. That would be nice. Not vital, but nice. And the other thing is that we're just going to try to write. So you would like to have a way to link with an anchor deep into a node or? Yes, which anchor that's not there already. An aposterior anchor. Like a annotation. Web annotation, exactly. Yeah, I mean, I suppose I would be like to reuse the annotations. Yeah, that's fine. And the second thing is those anchors are occurrences of concepts. And the concepts have nothing to do with how the anchor text. And we need a many to many relationship between anchor text and concept identifiers because people don't use words consistently. And it must be possible to say, this word in this thing is used the way that that other word in that other thing is. So we need to have a kind of translation table between occurrences and. Since the guardians of decency in Iran are about to be unemployed, maybe we could turn them into the word police so that everybody, they can enforce that everybody actually means the same thing by the same words. That seems like a good re-employment opportunity. Yeah. With even less hope of it even ever vaguely applicable. Good point. But yes. So yes, I mean, this seems like, and you have been in front of these, you say, hyper-knowledge mass is your. Oh, you're just opening the Pandora box there. You're opening. Yeah, yeah, yeah. It's a conversation. It's a conversation. But basically what I'm trying to build is what are the pieces needed exactly for that, for interoperability between tools for thoughts or tools in general, so that we can create the interesting links between applications. And so as an aside, I guess, I mean, personally also to say like this, sort of like high level mapping of what are the approaches that are trying to do this interlinking? Yes. That's the second map now we need, I guess. Yes, we need the project of mappings. Yes, absolutely. The mapping of mappings. Yes. The end, and one of the things preconditions is as much deep linking as we can, of course. And what we want is deep backlinks. But we can't rely on the anchor text for the meaning. So if we want to link not just text to text, but meaning to meaning, we need to have a kind of very supple relationship, many to many relationship between anchor text and meaning, and a way to negotiate meaning to say, I think this means that. No, I think it means that, and a kind of meta conversation about meaning. And when you say meta conversation, you're not speaking about a graph anymore, but a recursive hypergraph, because you need to be able to qualify each relationship. So this is what I'm trying to specify. And I know others are doing that. Yes. I am also currently trying to do that. Is anybody else trying to do this in this hall? Well, I'm working with Jack, so that's. Jack, you are also doing this. Nice to meet you, Jack, and I think also Peter, by the way. Yes, nice to meet you. Hey, nice to meet you, likewise. Hey, this is the first time I've seen you face-to-face, but I certainly follow you on Twitter and all of your work. Likewise, likewise. And Slutkin, if you want to talk a little bit about Agora, yeah, if that would be helpful. Talk about your approach. I'll talk more. I can talk eight for hours about mine, but. Yeah, no, no. I think we are on the same page there, and like, I think this is maybe the occasion, I mean, at least with the right frame and translation layers, where we could all do this for hours, because I have the impression we will be talking about like convergent projects. So in this framing, at least very high level of like, you know, recursive hypergraphs, we probably like, or I'll talk about the same approach. But yes, so the Agora is also approaching this space. And I started it precisely to do just interlinking, iteration, I call it. So opportunity iteration in the sense that, and this is also what I was saying, it is very optimistic when making links and we're actually like trying to integrate many different data sources into like a relevant but only occasionally cohesive whole. So this means an Agora is a device, you set up and it's like, you point it to a number of repositories, I will just consume them all and try to surface all the resources he has identified in any semantic context in which they could be relevant. So to sort of say this is sort of like an integrator plus some very basic search procedures. And yes, and it takes all these more nuance, like perhaps sub-problems, like how do you manage this meaning to me relationship between, it takes a meaning, it just completely like deligates that or like doesn't tackle that by saying, same text, same, like assumes a meaning and even originally just like doesn't like even listen to that in some sense. And you know, like tags, you can use like a, kind of a case tag, you can use like weak links, you can have slugs and all the old map to the same sort of like node which will be this fuzzy like entity space. So essentially what it tries to do is like put every resource in many nodes and each node is like a description of an entity, like a unique context and then it serves whatever is right there and in a radius. So it's like a fuzzy this score in that sense. Yeah, you're not doing semantics basically. No, if you decided to point to question of semantics. Exactly, exactly. With the hypothesis, convenient hypothesis, that if we get enough users, we will have a emergent meaning and a version of vengeance. So it tries to be none of you made it there. We may get a dumpster fight instead. But if anything else, it should be fun if we get there. And in the sense that, you know, like the rules what they also try to offer is a series of mechanisms to use the same basic building blocks the resources which are like text and links essentially is based on plain text markdown from formatting and weak links. It adds some directives is like a DSL sort of like a set of languages which can be described with weak links. So you can use them in any tool and those are like self-referential. So you can, so the community can use those to hint to say for example, I actually when I say this, I mean this but it's sort of like an optional layer. How do you say when you say this you mean this when you only have the internal links of the text? Like if I cannot edit the text, how can I add a link? I mean, that's my problem with a lot of markdown based tools. You don't have offline annotations so you cannot come from the outside and say I think this means this or I'm inferring this implicit type of thesis there or stuff like that. So right now I'm experimenting with like two procedures and I don't know if this is an answer but it is the answer that I adopted which is like two of the directives that I'm experimenting with are I call it pull and push just to try to use this like very physical sort of alignment which essentially pull will be like transcription. So that entity over there, take it here I bring it here because it's relevant in some way. And push is remote transcription in the other direction which means it's like a PASA action which means whatever I write here is also relevant there. And the hypothesis that you know like this is like service hints, strong hints that that remote entity is relevant here and you know, whenever you have like a symmetry of that where you know like I used to saying this around here and this around here that actually brings the entities together. And then that plus perhaps some definition of the radios you know will mean that you know if enough people are essentially referencing cross-references and not only that but transcruiting entities then that could mean that they are the same in some context. So that's yeah. Yeah, so you're trying to do basically those are I've been referring late in the latest FJB call to what I'm calling deep backlinks. And I think this describes quite well what you're doing, right? I don't know what deep backlinks are but that sounds to be the thing. Well, deep linking is linking to an anchor with annotation and what you want is backlinks on that or back deep links maybe I should say but I think deep backlinks sounds better. But it's still very much on the backlink model in the sense that the referred to entity has to be aware of the protocol to know well, this is what's being said about this. Right. Yes, there needs to be some like a receiver essentially. Yes, yes, yes. Yes, I mean, I mean, I want to read more on deep backlinks so actually I'm sure that I got this. But yes, I mean in practice, I'm not sure that the push and pull that would actually are enough to get those all these problems but they're essentially like, I'm trying to like add as few concepts as possible over just links, essentially, which is what Marlon has. No, yeah, and also with the idea that, of course the idea is that all of these work no matter in which tool you use. And in particular I was tackling a different role in the superhouse, the superhouse is like the issue of like block references, you know, which are like presenting many tools and they actually make it sort of hard because they depend on like a centralized sort of like registry, right? Yep. So in this case, essentially, it all uses a local context so that it can be anonymous to some way. And if you know, so you could push and pull from no written, you could push and pull blocks written in different tools. So that's the pitch. Right, right, right, right. Yeah. The, yeah, I think I'm, I see a lot in common, but I see a lot that is different. Let me focus on the differences though. I think there is a lot in common. Deep back linking is certainly something I want to do, but this is a, this links, so to speak, occurrences, right? And I'm thinking, how can we think at, how can we do collective intelligence at web scale? And that means any single idea will be in millions, hundreds of millions of documents. So the link tool will, I think, become unusable at that scale. And so I'm interested in if we can identify concepts and de-doplicate concepts, then we're able to say, okay, this is this concept in this document. Here are adjacent concept like distinctions, argument claims, and those claims would be each have, you know, maybe a thousand documents where they're made, discussed, blah, blah, blah. But if you have some, the point of having semantic identity as opposed to syntactic is to give a less overwhelming list of references and also be able to focus on some social aspects. Like it's not important to know that a lot of people think that dogs are nice because a lot of people agree on that. It's not contentious. Whereas the dogs are dangerous is much more contentious. And it's that way, knowing who aren't used on either side of that is much more interesting than the really uncontentious, uncontested notion. Of, for example, dogs being crudder beds. Like, nobody's bothering about that. And so having a kind of, what's the social truth about this? And I'm really interested in the fractal nature of this. Like this is what I believe about a statement. The statement, I believe to be true, false, plausible, maybe interesting, I don't know. And that's one, you know, it's not just true, false. For me, alone. And if I'm in a team, then the team may say, we collectively believe that according to whatever social decision mechanism we have in that team. Majority, leadership, whatever. Whatever. And then socially, then this is a numbers game. Like, what do people believe and why? More important, why? And can we get people to go from a commonly held belief to why do we actually believe that? And expose that very prominently. And maybe other simple entry points to that exercise. So I'm reminded here of Dave Gray's diagnostic card deck for organizations where he wrote down a whole series of sort of dysfunctions and high functions that organizations have and he would hand them to a potential client and say, here sort these into a pile of the ones that you have in your organizations and the pile that don't apply. And it was this like brilliant little thing where it would collect down to, oh, these are the things we should be talking about. And here I'm thinking, if we want to figure out people's belief systems, is there something as simple as a virtual card deck, like, you know, flip right, slide left. The tinder of cognitive biases and... The tinder of internet hatred or whatever. Like, yeah, swipe left, swipe right on whether you agree with this particular statement can rapidly get you into some constellations of belief and conversation. Interesting. Certainly the point is to go beyond the swipe, right, swipe left into why do you believe that? But I agree that having a quick picture of how would you believe is the entry point to the next conversation. Right. So, yes, that seems really the ultimate goal we have here, which is like a really quality of discourse and your integration coordination. So, I think from the point of view of what my very focus in all is to be more on the mechanics of integration. Yes, it is completely open but... Occurrence relations rather than concept as pivots. Exactly. Which is core for me. Yes, although I will say that the idea of pull and push, for example, when you pull an entity is that the entities are pushing the other one will also be pulled. So essentially it's bringing, it's actually bringing close to like a section of the hardware graph. And in that sense, it also has, of course, the social aspect, which is graph is contributed by different users where like the idea which is not meant to be able to say, okay, show me, let me browse the graph like I was this user. And I'm not looking for... No, that's important as well, yes. I'm thinking of every user having one or maybe even many event streams of here are things I believe and here are. And then using that to have a multi-perspective and be able to make cross-perspective assertions like this user thinks like this one on this topic. So, my, sorry, go ahead. I'm gonna throw a pragmatic and impractical real-time real-world example into this just because it seems really important, but there's a problem here that half of the Republicans running for office in the midterm in the U.S. right now believe that the election was stolen. They're basically election deniers. Or say they believe. That's fine, but I'm really interested in it. Is there some sort of way to knock them out of the ability to participate in civic discourse if they agree to that statement? So, do you agree that the election was stolen could become, in my dream world, could become a litmus test for whether or not you get to participate in the state legislature or whatever else. That denialism of election and therefore the undermining of the next elections is a reason to say you cannot in fact show up to talk in this forum. Because otherwise we're screwed. Like a whole bunch of these people are going to be elected because they're in safe districts and our legislative bodies are going to be infected with people who believe this as from the get-go and nobody's taken any kind of draconian measure to make sure that doesn't happen. So I'm kind of proposing a draconian measure based on the conversation we just had about a stream of what people believe. One thing I have been advocating for quite a while is if we're going to go into collective decisions we want to qualify believe not just by a number of people who believe it but by their track record of being aligned with reality and especially in the predictions like there's been all this magnificent work of super forecasters, yes I know I keep repeating it. Great, I bet he's going there, I bet he's going there. And you are right, that luck has done wonderful work making it very clear how to measure the ability of someone to align with reality even with fuzzy predictions. That's really interesting. This is also what the Dot Connector does at Dalio's Ridgewater. So one of the reasons they rate everybody really openly and kind of it feels like maybe too transparently and viciously is that they want to figure out exactly what everybody's track record was. So if Bob across the table is always annoying but always right, then we should like get over the annoying parts of how he delivers this message and listen more carefully to what Bob says next time. And the only way to do that is to track the assertions made over time and then figure out how to rate them. Absolutely fundamental, and this is why the track record matters and being able to, and that's the other reason I'm text, text is what it is, but getting conceptual also is a way to distinguish empty claims from testable claims. An example we've been working a lot with as a test case in CDL was someone saying the economic cost of the lockdown is greater than the health cost of the pandemic, for example. And this is a perfectly meaningless statement in the sense that you can imagine totally different scenarios with totally different costs of either health or lockdown, economic costs, health costs, la la la, and where the person could say, yeah, that's what I said, I think it's too high and with totally different scenarios. I mean, if somebody says that, okay, what's your prediction for the economic cost, what's your prediction for the, and Bob Ranges, fine, but then what's your prediction for the cost and what's your conversion factor and what do you actually mean? Because a test has no business example of a non-testable prediction that you'll find in text. And until you actually try to disambigot and say, hey, what is the meaning here? You won't know it. And this is why the mapping of language to meaning, which is a specialized activity. I don't think we can ask everybody to do that, but it's definitely a crowdsourcable task that can be taught, that can be shared. I wonder, so going back to the three of the problems we have is because, you know, like, for example, one possible situation we are in is that there's enough people in the, say, no day came, weekie collaboration spaces to actually perform this task, which is socially very meaningful, but they are not permitted enough, which seems to be the case, it just can't happen. So then I guess my question will be, what is the minimum set of actions that we can take so that, say, this is due to the crowd, right? Cool, actually, for example, like on a date, you know, and qualify statements done in, you know, public day settings. Like, one of the first problems I can see, which will solve is how do you refer to an event in such a way that, you know, do you need at least that coordination available? You need to be able to say, and this is where, like, I sort of feel that having one way, you know, like, it will be like typical thing, like, what is the hashtag for such an event? So then we get the middle part in following me, people use hashtags, may it's probably like, you know, it'll be like encourage. You encourage, yeah, thank you, thank you, yeah. And this is where I'm of the strong opinion that hashtags are not enough because they're too ambiguous. And what I want is the capacity to take this piece of text and say, this is a more formal meaning of this, or this is what I think this mean, this word means in this context, and then crowdsource that, and then say, okay, people are disagreeing on this one, so maybe it's an ambiguous word, maybe it's a dog whistle, if there's disagreement on it, and note that. But the reason where I guess I go back to it, for me, the minimum coordination action is, I agree to annotate like, each event in such a way so that it can be later recovered that I meant that event. You know, in my approach, and my notes are like all over the place, sort of like, because that's what I can do, and this is a great big point, but. But you are in the Acora, right? Part of it is, do you have a centralized place where all that data is kept? So the, which is, you know, both good and bad. So, well, we call it federation in our world. No, I'm well aware of how that works, but even in that space. So you've got, and you could look at it the same way, you could look at the long tail of hashtags on Instagram. So to reach an audience, an Instagrammer will add 57 hashtags to the end of their posts in hopes of casting the widest net possible. And usually it's variations of six different words up and down and left and right. But if you're Instagram, you can say, okay, you know, this post is essentially got at least one hashtag that says Academy Awards is an example. And then it may have seven actors' names and four movie names and something else. But that massive high level aggregation can look at that even 20 or 30 people who are doing all those hashtags to aggregate them and then tease that out. Or you've got the same issue, you know, let's say the Super Bowl is on in February, you'll have a million people on Twitter talking about it, typically with no hashtags saying that was a good play or bad play. And the only way to recontextualize it is to find the data timestamps that relate to a larger world map. But Twitter ostensibly could do that and add an invisible hashtag of that metadata to indicate there's a high likelihood that this thing is about that. But you have to do it from the top down to be able to recreate those contexts in reverse and then add a probability score of how likely it was that they actually were about that particular topic, given who the person is and their past history. So you can do it with some ergodic theory, probably, and actually tie those things back. So it's doable, but it requires having some central or even federated repository of data that you can access to reverse engineer it. And it shocks me that we don't do that. I mean, Instagram would be a much nicer place to visit if I didn't have to see 500 hashtags on everything influencer, why, posts. Yeah, makes sense. So just quickly, the idea behind the social take we're taking or the federated database that the community agrees to be in one hour, sort of is agreeing to cross link promiscuously, to some extent in the sense that, if I link something like an entity to link, then that sort of gives, it gives the algorithm that that same thing could be linked in the same way, even if it is not linked elsewhere. So to make that this like a some forceful community linking. So to make that this both the direction of like converge and like actually be able to have this higher level, like a group of hashtags or links and high level like your semantic linking happening like consensually by a community, right? The, yeah, my concern is whether we're speaking about real hashtags or computed hashtags, we're still at risk of confusion. And it's not just the ambiguity. It's really the, what do you mean by that? And being able to have that question and that question being made kind of salient to a rate. So we're really can say, no, no, I mean this one, I mean that one and creating the distinction when it's meaningful, when it's important so that everybody can reuse that distinction and reuse the ambiguity. And if we go back to political discourse when somebody uses a dog whistle, ask, do you mean that? And so the politician kind of has to say, well, no, not really, which is probably a lie. The problem is no one's feet are actually held to the fire. You can say it was that a dog whistle and they'll deny it when in fact it was and but it can be shown that a lot of people understood it this way. Well, this, which is already an interesting thing. So they'll have to deny it more strongly. A lot of people said, I understood this. I read this as a dog whistle. And I mean, it's this last week's LA city council, Bruja with multiple people either resigning or being asked to resign is, you know, where it is latest, you know, so a group of ostensibly diverse city council members were having a conversation roughly a year ago and used a few kind of racial code words that exist within the Latino community and an audio tape was released. And it became a big enough thing that Joe Biden weighed in and said, yes, you should resign from, you know, the president of the city council. And this is in a very left-leaning area with generally progressive people. And like, and likely next mayor, I think, too. They were being, you know, broadly racist in their statements. And it becomes a thing that they would only say that or they said it because it was a quiet closed group that they didn't expect it would get out. So there's always that. And right now, I think that's one of the bigger issues with most of the Republican party is they do as I do or do as I say, not as I do. So they're all pushing out messages and then doing the polar opposite. And that becomes even a harder thing to fight against because even if you say something and you clarify it and you're super clear, the fact that you can go do something totally different and you're judged by words, but then your actions are wholly different is a whole nother level of problem. That is absolutely true. The, I've been having a lot of argument with people who are deeper than I am in the Dow and smart contracts days. And I'm totally, I mean, I think the Dow crowd is fantastic, but I'm really not comfortable with smart contracts. I think they are just opportunities for bugs and cascade effects. But that said, the good thing about smart contracts is enforce accountability. It's having forcing people to act according to what they say they will do. And there is something there. Despite my discomfort, I may need to the question of how aligned are words and actions? When words are legal and have legal value. Small side note in the free Jerry's brain calls some months ago, David Baubill, an occasional member walked us through Lexon, which is sort of a language for smart contracts that's meant to be a human readable and machine readable and more precise language than the normal garbage you find in contracts. It was super interesting where you could sort of, you could sort of see something important coming together in that space. That last intervention wasn't very much influenced by David Baubill. Go ahead, Stensin. No, I want to do a very clarity on the name, is it Lexon? Yes. Okay, thank you. Right, makes sense. So I guess, well, this is a very complicated topic. I think, of course, like, if we're going to be in the direction of how to fit the political system, which I think we should do, by the way, but that seems like it carries a lot of complexity. I wonder if we could perhaps think together about what the prototype will look like, that let's say a community do something in concrete, like better when it comes to pushing back against, like, just complete households or this like. Yeah, I have a prototype for kind of the bigger vision, not the prototype, sorry. I have a plan for prototype. I'm working on the data model a lot and I've been working on it forever. And my God, it's slow, but it's progressing. But I have a roadmap for what I call hyper knowledge, which is the unifying thing. But it's true that we could have useful things shorter run for the political side of this. And that's totally a good question to ask. I'll have to think hard about what's a good subset. Even on the political side, I think something that could help would be watching the ability for people to micro target and make their actions and words not match. So it's very easy for Donald Trump to micro target people on Facebook. Who are anti-Semitic because you can do a search and then send them specific mail. And if that's the one issue on which they're gonna vote, and that becomes the majority of their decision, how do you then, and essentially it's a hot mic moment, but it's a hot mic moment, no one is going to say anything about because it's a quiet like, oh, I need to be signaled to that this is the case. And the way Donald Trump speaks, it's a very, you can look at old things like techno battle or double speak or political speak as examples. He's saying things, but there's not enough semantic meaning and any of the things he's saying because they're not full sentences that literally anybody who's hearing it can draw the perspective of agreement that matches with their particular polarization. So early on it caught a lot of people both in the center, right, the center and even some on the left. Oh, I could be for Donald Trump because I hear him saying these things. But at the end of the day, none of the things he was saying actually meant anything because of the way he speaks. And in his case, it's a very linguistic, difficult thing to pull out. So then you can only act on, and it's been shifted because his actions then became totally apparent what his policies were. And then he lost that fuzzy center, left center and a lot of the center, right? Because it was,