00:00:00 Speaker: Sick. Wow, this is awesome. So many people. What's up Padawan? No, they're not here. Um, good to have everybody. Thank you all for coming. So we have Kylan, who seems like he's trying to work on audio settings. Not a problem there. Um, but just to kind of set the stage, you know, I can't say that I'm the absolute best in going to have all of the, all of the answers here. Um, but when, when creating data, I guess this call is just for a number of things. It's a way for us all as builders to connect with each other, potentially show some demos to each other about what you've been building on, new features, different things of the sort. Um, also like if you're hitting any roadblocks in building, this is a great time to surface those. Talk with each other, get through, get through any roadblocks and hurdles that you might have come across. But then ultimately, a very big part of this is something that Kylan has brought up quite a bit. I know the core team has been working on and chatting on as well internally, um, coming up with like conventions around claims and predicates. Uh, yes. Kylan. I am recording in a couple different ways. And um, so anyways, I'll see if I can get this fixed. Uh, would you be able to make it to where this is not pushed to talk? Any chance that this channel can be changed to where everyone can just speak by chance? Yeah. Um, and so ultimately going back to what I was saying, uh, in creating claims, there's going to be certain claim structure that we all want to use as builders that are essentially a standard for how certain claims are made. Coming coming to convention on which predicates are used and which, um, in which event and things along this nature, which are a bit more technical than my head can handle. There are things that I know Kylan is great at thinking about. Is that as well? So not so much here to lead as I am here to facilitate and just make sure that things move forward and whatnot. Um, that said, Kylan, it looks like we can hear from you. What's up brother? What what do you have to to add to my little bit of intro for what this call is? You can you can or cannot hear me. We can hear you. Yes. Okay, good. Every time I do discord, uh, for conference calls. I'm just not super used to it. You know, Google Meet and Zoom. I'm more used to. Um, but yeah. So, uh, I guess, do you want me to take it from here or. We're, uh. Yeah, man. Um, absolutely. If you have, if you have ideas, I know, like if you have ideas for how this should run or specific questions because I know you've already kind of been thinking about these things, right? So I'm happy for you to like run with things if, if there's anything I can do to help. Of course, anyone else here with us, feel free to chime in and give your opinions, but happy to hear how this goes in your head as well. Kylan. Yeah, so this is open to anyone who wants to discuss anything or do a demo. Uh, if you don't mind, I would like to do my demo and kind of go through my flow. Uh, it's obviously very, very high priority for me. Um, uh, let me see here, let me share. Uh, that's what's up. Yeah, we'll get straight into it. I know you guys like demos. Everyone likes demos. So yeah. Every time I press entire screen for sharing my screen, I'm like, ooh, I really want to do that. I know that feeling. Yeah. How sure am I? Um, yeah. So you guys can see my screen. All right. Yes, sir. Okay, so this is the ugliest slideshow. Uh, so let me just say that the focus for this demo is the demo. And to get feedback, um, the actual visuals for the presentation part I just scraped together. Um, I essentially what I've been trying to do is trying for the demo I'm doing, I'm trying to get it so that we pretty much have the whole flow. So you can kind of see how it'll go. Uh, the tricky thing is the flow, you know, up until the last minute, I'm working on connecting it. So there's probably going to be some bugs. It's going to be a little spotty, but that's not the end of the world. Uh, so for those of you who don't know who I am, I'm Kyle and I go by smiling Kyle on a lot of my usernames. I've been building web apps for about God fifteen years, fourteen, fifteen years now. Uh, started working full time in crypto about twenty seventeen. And then I started working. I found out about intuition through Billy. He's giving a presentation at consensus. I was working at MetaMask at the time. Uh, that was in August twenty twenty four. It's crazy to think that it's been that long, but it has been quite a while. Um, so I hive mind is essentially it's mostly me, but I also have a designer who I'm contracting and also we're working on some animated videos. Uh, but with regards to like actual employees, full time people, that's just me. Uh, right now we're working on Multiple intuition products, and the browser extension is going to be the main one that we're going to focus on today. We're also going to show the Intuition Explorer that's going to be part of the demo today. The MetaMask snap I've done demos of before. We're not going to really be showcasing that today. That's in a good but unfinished state. Um, we'll be talking about why in a moment. We also have a marketing website, hive mind HQ dot io. It's not the prettiest right now because we're not really focusing on that until we have our actual products ready at that point, once we have them ready, then we can make the website look prettier. Although you can go to hivemind HQ dot io slash curators if you want early access. We're going to be talking a lot about that today. So the purpose of this demo specifically, like I said, this demo is not well polished. Um, I'm not going to say that people can't share videos of the demo, but if you do share any videos, keep in mind how it makes our community look how it makes my products and my company look. Um, so if you want to share it, I'd prefer if you picked a clip that doesn't look terrible and also maybe add a little bit of context wherever you share it. Um, so the flow that we're mostly going to be looking at today, there's a few parts. First one is the flow for a new user, uh, to the hive mind browser extension. Then we're also going to be looking at what it looks like when a user creates a new topic on intuition. And also when a user decides that they want to become a curator. Um, so let's see here. I'm going to skip down here real quick. So what we're doing, and I think most of you guys probably already understand this, but originally I thought we might be able to do more of like a looking at global claims on intuition and aggregating them, looking at statistics and trying to sort out the noise and find the signal, so to speak. And we found that really difficult. Um, it's, it's, there's a lot of noise and I'm not saying you can't get through it with all the signal, but you're talking about proprietary algorithms. And in a web in a space like blockchain technology. Uh, if, if users, if there's algorithms and users can't verify, it tends not to work well. So we've had a lot of discussions, especially on the message boards about the idea of trust circles. So I'm not tied. I'm not married to this, uh, triple structure, but I think that's what I'm going with. So essentially, if you want to trust someone for a topic, you're going to have, okay, their address. And then the predicate we're using in the middle is trusted for. We feel like that ties the address to the topic. It's very concise still. And then the topic itself. We can have a debate some other time about, uh. Um, I'll probably make a post after this presentation. Um, but like, I like to use simple text, uh, atoms whenever possible. I don't like to use the IPFs ones. I know a lot of people like IPFs. Um, but yeah, we can discuss that in detail. I don't want to get too bogged down in the details just yet though. Um, let's see here. Oh, another thing this is going to be a little bit cool is you're going to be looking at some of the user experience improvements that I've been working on, uh, to try to cut down, especially how many transactions we need when a user is making a claim or they're staking on a claim, uh, if someone has to create the atom, then create the triple, you know, there's, it quickly becomes a lot of transactions. So anything we can do to streamline that, especially when we're trying to build up the network graph, is critical. So what is this demo. Not again. So this is this is not a finished product. This is me getting an idea of what you guys think about kind of this, this the flow and the structure. That means that style decisions here and there. Uh, we shouldn't that's not the focus of this demo. Um, a lot of I have a designer, but he has not had the time to touch a lot of parts of the extension, especially these parts that I've been working on. Most recently. The icons you're going to see, icons are going to play a critical role in this. The icons I'm using are kind of default icons. They're not, they're not pretty. To put it simply, uh, the, the web app, the explorer web app, uh, we'll go over that a little bit, but that's not really the focus. We're more focusing on the extension. Uh, we're also not looking for any bugs here. You don't need to point out too many bugs. There will probably be some bugs along the way. That's just the nature of it. This isn't polished. It's not ready for, uh, for you for public release. Uh, some of the statistics, like, uh, the amounts might be slightly off here and there. I don't know if that will really happen, but I'm not here to sweat that sort of stuff right now. We need to make sure it works. Then I, we can make sure all the all the data is one hundred percent accurate later. Um, yeah, there's also some onboarding extension flow. You'll see a little bit of that. Uh, that's okay. We don't need to focus on that. Um, okay, so let me think about this. If I'm starting off, I will start off making a topic. So if we have a, a new user and they see this, they, oh, hey, I can become a curator. Oh, by the way, do you guys all hear me? All right? Everything's good. Yeah. Still clear? Yeah. Still clear from my end. Okay, cool. So let's see here. So if we go back here. So I'm starting to put out calls for curators, and we'll go over what that means in a moment. Essentially, you're kind of going to be going to be like an influencer in the intuition space. You are going to, uh, here, I'll show you. Um, you're essentially going to assert yourself as a curator and people will be able to trust you for a specific topic. So that means they're adding you to their trust circle for that topic. And then I'm going to be demoing kind of what that's going to look like. Um, so if we have a user. So this is the hive mind, uh, explore web application. Uh, some of this is not that important right now. This trending stuff is kind of a lot of noise, but if we go and I'm zero x e four, five, whatever it is, and I go to my profile if I want to say, hey, I am an expert in something. What you got to do is you get to declare it here. Um, we have some whitelisted ones, I think. I don't know if we're going to keep those up there, but you can choose different topics. So like if I wanted to become, uh, so you'll see there's some visual bugs here. Uh, let's see here. I think sports or if I want to do tech, we'll do tech. So if I wanted to become an expert, a curator in tech, click that and then this is going to declare your expertise. Let's see here. So it's just one transaction. It's a claim. So what happens there is we're using that same structure. Was it trusted for. So in this case you are trusting yourself for the topic so that we take as a convention saying that, hey, I'm asserting myself as a curator for this specific topic. We could use a different triple. Uh, it's not a huge deal. It's, uh, arbitrary at this point. Kind of. Okay. So we, we declared ourself as an expert for tech. Um, I actually don't know if the topic is going to show up right now. I think there has been a delay sometimes. Um, but let me show you what it looks like. Then when you are so this is what a topic page looks like. Okay, so when you create a topic, so like this is what we did for crypto. It's only a topic because someone said that they are an expert for it. So one thing that we need that's critical is when you have users coming in, when you have curators, we have users coming in and you want them to add people to their trust circle. What you really need is you need multiple curators. That'll help make sure that there's a lot more coverage. Um, so we need more curators. And also, so we're using this, this screen as a bit of like a health check for the topic, especially early on. So this is early on, curators were saying, hey, try to get to five right now you have one. One curator for topic isn't healthy followers. Figure ten is about a good a good number. If you have a lot of curators and no one's following them, then you know, does anyone really care? Claims this week I actually think this is incorrect. Uh, so don't mind that. Uh, this is kind of cool too, if someone's browsing the topic. Oh, do you want to be one of the first curators? Or if you want to get other people to, to add themselves as a curator, right? To help improve the health of your topic, then you can have this invite link, which will, I think will bring them to this page or something like that. Um, that way, you know, if you want the topic to thrive, then we need to kind of do whatever we can to drive users to. Um, to get more curators. All right, so let me think here what the next. Okay. So let's go ahead and pretend that we are a new user in the extension. So I'm actually going to start from scratch scratch here. So if I'm starting, so this onboarding flow is it's not one hundred percent what I want, but I'll give a little bit of a preview just so you see the functionality when I install this app. Should do this. Sometimes it takes a while. There we go. So I'm on in developer mode, so it's not as fast. Sometimes I think first time around it has to parse a bunch of Json files or something. So this is kind of the onboarding, onboarding flow. I'm trying to make it as interactive as possible. So this is for any user who's adding the browser extension. Okay, so if the community thinks about anyone on x.com, the extension icon I con, you don't have to pin the extension icon, but if you want the best functionality, you're going to want to do that. It's cool when you do that. It can tell you did that and you're like, okay, well what now? What am I supposed to do? We got a little thing here that says, okay, here you go. So this, this sidebar comes out, this is the hive mind browser extension. Okay, cool. We'll see it in action. Let's go ahead and do that. Okay. Glowing badge something something. Okay. So if we look here. That's one that makes sense. If you're using hive mind, you should know what intuition is. And we want you to becoming familiar with it. Okay, let's click on this. And this is cool. So you can see, you can see people have made. Go ahead with someone saying something. That sounds. No, I think that was just background noise. Okay. That's fine. Uh, so yeah, so you can see these little badges, you click on them, it'll show you claims that people have made specifically about that x dot com account. Um, it also tells you okay, weigh in. So I don't know how prominent this part is going to be. Early on when we were trying to match people with curators, we were giving them kind of pre-approved claims to vote on. I don't know if we're going to be doing that going forward. Uh, so we'll skip to the next part. Uh, this is weighing in on crypto. Okay. Go to trust. Um, trust matches. Okay, so here's where it's going to show you people that you can trust. If you would like to trust someone for a topic right now, it says you can't because you're not logged in. So let's go ahead and connect. I want a different. Let's see here. Okay. Let me pick a completely different. Okay. So if this person is. Okay, so I'm going to use a, a kind of pre-approved list I made just to make this better. So if I go to list, because we're going to be focusing on the crypto topic specifically. So I made a little bit of a list where it's like some, some relevant accounts. Uh, let's see here. I'll make it bigger. Oh, I don't want this yet. A different account. We'll do. Yes. All right. So if this person is browsing around. Here. Okay, I'll make this bigger. So if you're browsing, you can see that there are little badges here. So this says if I click this, it tells me I'll get out of this stuff. A bunch of onboarding stuff. Um, so if I click on that, it'll bring up CoinDesk. You can see the claims that other people have made about it. You can also make claims yourself. Trustworthy builder. Right now we've been kind of using AI to come up with like, okay, well, what's the best, uh, what are the best claims to make for a specific type of atom? Like, is it related to a topic? So that's a lot of the inference that we kind of have to do either through our own logic in our app or through actual actually using AI. Let's see here. So we've got CoinDesk. Another thing too is if I want to make, uh, claims. In fact, let me make this more compelling for you. So this is I'm going to give myself some trust tokens. So that should go now. So now we have a balance. Um oh, so this is kind of cool. So let's say this user wants to start making claims they can. And these are kind of like quick, quick action icons. Um, so if I click some of these have meetings. Okay, so these icons are not the prettiest. These are the default ones I was talking about. So if we go in this direction, which I think we will, we'll be coming up with probably a lot of our own icons. That'll be a little bit easier to figure out what they mean. So you can also imagine if this person happened to be a curator. I can click this repeatedly and you can you can't read there, but it says six trusts. And then every time I click one, it goes up. If I go for the same claim that it just adds more on. And I also have it so it sets like a default. Let's see here. Not set will prompt on first. Okay, cool. Yeah. So if I keep going, let's say, oh, I'll say I want to say that their alpha provider coin market cap. Same thing. Alpha provider Cointelegraph. I'll say that they're trustworthy. So now I have four claims in here. So I've cued my claims. So now what it's doing here it's preparing them. So it's saying hey, do these atoms even exist? Do the predicate atoms exist? Do the subject atoms exist? That way when we actually do submit for claims, it's just depositing on four different claims. So if we do this. There we go. It's supposed to. I'm supposed to have it ask me how much I want to stake per claim. But that's okay. That's not that critical. But you'll notice. Did you see what just happened? Now these gold icons show up. Okay, so yeah, let me see if there's any. So the gold icons mean that we have added a claim of our own. So that's great. And then, uh, I'll go ahead and clear this out for a second. So the other thing is we haven't trusted anyone yet. The problem is if you make your own claims, you don't care that much about your own claims. You already know what you think of. Okay? You already know what you think of CoinGecko and on Cointelegraph. Um, I mean, this is kind of cool. You rate it as trustworthy. Cool. Uh, so if I go here, so what it does here is it looks for curators, um, we're still working on the process for signaling or surfacing the best curators for someone. That's an area that there's going to be a lot of work on. I think we have a lot of ideas that work really well. Um, early on, it's trickier. So there's worse comes to worse if we need to, we can have kind of starter packs or something. But once you get a few curators, which is why we push for it to, to have multiple curators is then if you, if two people say they're curators for the same topic, then if they are making a bunch of claims, the chances are anything that they make duplicate claims on where they have overlap that those atoms Are probably going to be more related to that specific topic. This does beg the question of, okay, can a curator be a, a, uh, a curator, an expert for multiple topics? I haven't gotten quite that far. Um, but you know, we'll get there when we get there. I don't, I don't see that being too much of an issue. Uh, and it does make sense because some people will be experts of multiple topics and sometimes those topics will be related. Right? One might be crypto, another might be, um, secure, uh, smart contract security, for example. So I'm going to try something here and hopefully this works. So if I were to click smiling Kylan so you'll notice the feed doesn't have too much under the avatars. This is one gold badge that we did ourselves, but generally there's not a ton. So now if I were to add smiling, Kylan, who is a curator for crypto, as you can see, crypto, I nominated myself. Uh, at some point, I don't, I don't know if I'll consider myself an actual, uh, curator for it, but for the point of this demo. So what we're doing is we've trusted smiling dot eth to our trust circle for crypto. So I've added him, um, and then now supposed to do a live update in here, but I don't know, sometimes it takes a little bit while of a while to, to update. Yeah. Looks like maybe it's a bug. Oh, no. There you go. See, it takes a little while. So now you see, okay, these green icons, like it's trustworthy, I think is the one here. Yeah. And you see the avatars for the people from your trust circle who specifically have endorsed that claim. So I think that is the main part of my demo. Uh, I guess I'll go ahead and stop here. And just what are your guys thoughts? Anyone got feedback? I mean, I'll just say from, from my own point of view, I, I think it's really cool. I like how you, um, I like how you've integrated it into the, the Twitter feed to be just like right there under the PFP. It makes a lot of sense. It seems very easy, like very intuitive, very easy for me to, for me to use click through, make claims. I think that's one of the biggest things from a user point of view is like just making things as easy as possible is huge. And you've done great with that here from what I'm seeing. Yeah. And one thing I'll show you real quick, this is one of the things that I really like about this. Um, if I close that sidebar because most people don't browse with, with a browser extension sidebar open like that. But so even if I close the browser extension, if I start scrolling down, I will still see these icons. So even if you install it and add some people your trust circle and forget and never open it, you're still going to see, uh, the, uh, the insights from intuition. Um, and as a side note, it's a challenge sometimes to get this, these elements, these components and icons into this little area. Actually, this area is okay, but in general, building on top of other web sites can be a challenge. I do think this is a pattern that we can do on other social media, not just social media. On other websites. So I thought, you know, maybe YouTube would be a place we could do something like that or farcaster cast or we can kind of, I'm not going to say trivially, but we can pretty quickly add other websites that we would like to support. Um, so yeah. I see, I see a good question. Have I thought about doing that for the entire web? Um, so originally what I was doing was if I were to go just to like google.com, then if we had any insights, uh, because in the background, the extension can know what website you're on if you give it permission to, we could let users make claims about the specific website itself, right? Um, then you could cover every website as a website. But then also if we go to, I'm not even going to go to YouTube. If I go to YouTube and we'll do it this way. Uh, if we go to YouTube. Let's see here. Click on something. Every kind of every social media platform kind of has, uh, some places where you could do it. Uh, so maybe we would do it right in here. Um, yeah. So anyways, you can find a spot pretty much for any, any website. Uh, this does beg the question, like, am I going to do that myself? What would I hire people, a bunch of people? Um, or would we make it open source or something? I'm kind of flexible on that. But yeah, this is something where you, it's, this doesn't need to just be x dot com. You can do this on pretty much any website. And ideally we would want it on every website. Yeah. Alright. What other questions we got? We got. Looks like lute is typing something out. Maybe he's got a question or some feedback or any other feedback from anybody. Let me make sure I haven't missed any functionality for the demo. I guess question from my end on like how you're querying for this stuff because it might help inform some of the API work. Like are you, are you, you're probably querying every tweet one by one, right? And then filtering by positions. HMM. Uh, so I don't do tweet by tweet. Uh, we scrape the browser extension does have access to this web page. So it will, it will look through. Check, check. Sorry, I heard echo. Um, so it will go through the each of these it'll scrape each batch. It's usually like six or eight of them at a time and it'll look at the usernames. Okay, so you're right. This is an important part. So it will look at the usernames. But the problem with usernames is that there they are mutable. So we have to use immutable IDs on x.com. Yes. The tricky thing there is in order to map them user IDs to usernames, you do need access to their API. Um, so that is, that is friction, especially if other people want to support, uh, these triples and these types of claims. Uh, I think if we, if this turns into something that is compelling and a lot of people use, I think most web apps and explorers would be willing to, to support the X username to, to ID, uh, functionality. I think it would be worth it. Uh, you can cash a lot of it. Um, trying to think what else? Yeah, yeah. So we'll go through those atoms and. Yeah. I think here you go through Adam's looks. It looks for if there are any claims, and then specifically, if there are any claims from anyone in any of the users trust circles. Nice. That's you have more available than I thought. So it's basically like you're kind of doing intuition queries and then a Twitter API behind the scenes just so you can map whatever's in the Dom to an actual user, right? Yeah. I've thought about sharing that map. The mapping of usernames to user IDs, X is somewhat strict about what you can and can't do. They don't want us doing that. So if someone wants that data, they have to query themselves. Uh, which is somewhat unfortunate. Yeah. So if, if, if, how do I say this? The actual atom label is literally in the format of like x colon one two three four five six whatever. Um. And that from a usability standpoint is not the best, right? So we do need to try our best to resolve those. We gotta put all those links in intuition so everyone can use it and make it cheaper for all of us. I mean, I, yeah, I'm trying to think here so you could do it automatically. And then that just means when it updates. Let me think about when someone changes their username. Then I would have to update the claim, but that's not that bad. And the thing is, I know some of you might be thinking, oh God, there's, you know, a billion people on on Twitter and yeah, there are, but realistically, like it's, it's a, what do they call it? Pareto coefficient. So it's going to be like a thousand accounts that are probably making up ninety nine point nine percent of the activity and following an engagement on X. So that has been very, very useful. So we use that to our advantage. Leslie, I'm less worried about like the amount of accounts because those IDs are immutable, right? So then but people can change like the cosmetic username. The username. Yeah. And they use the cosmetic username is what you usually want to show in like an actual UI that's public facing because X one two three, four, five, six doesn't mean anything, right? It's unless you've memorized which no one has. So. James, you were going to say something. Yeah. The thing that really caught my eye and that I liked, we need to think deeper about is like using these triples as search algorithms. You know, I was mentioning, uh, you know, we're obviously working on something and, you know, working on sorting and ranking. But I think having something public that really, in addition to this, like predicates that we've published is also how our search algorithms are built on, on triples. Um, I feel like that's the next progression of like, you know, these layers that we build. It's like, all right, these top one hundred predicates and then, okay, it's these like ten triples that you can build like search algorithms on. And I think his community, it makes a lot of sense just come to consensus and, you know, publish what triples you're using. We could probably even use those the same, uh, as we kind of surface different perspectives and easily finding those for people. So I just wanted to point out, I really like that idea of using triples in that way to, to start getting those, those search algorithms kind of, you know, published and documented. What do you mean? What do you mean when you say search algorithms? So you mentioned you were doing search like, hey, this person's trustworthy. Therefore you should, instead of finding every, uh, returning everyone's opinion on this. It's like this person I trust. Therefore, I'm going to surface this perspective sooner before others, right? So it's like affecting how your interface is behaving. We could adopt that, that same approach where if those triples are available, we surface that data to users ahead of time. So just influencing how these, these interfaces kind of respond. Yeah, yeah. Early on we I'm hearing here yet though I'm you. Thank you. Um, yeah, so early on, uh, again, like I said, I was thinking probably that global claims would have a little bit more, a little bit better signal to noise ratio and maybe it would. Okay. Um, but I know what threw me off is there was a contest that intuition was holding and the claim, the, the number of claims was something absurd, like one hundred and twenty thousand or twelve thousand. And it becomes very clear that someone's gaming the system. So it's like the second you make a statistics a scoreboard, then that statistic is no longer reliable because now it will be gamed, especially when there's staking and money and contests involved. So yeah, we're leaning really heavily into the trust circle functionality. I think that is the way to go. I think if you look at your life, you trust people for different things, right? The guy that you trust to fix your computer, you might not care for their relationship advice, you know, or their or their stock picks, right? Um, so the, there's a cold start issue though, because when the users join, there's not many claims. There's no trust circle, right? So you do have to show them something. We show them global claims. That's enough. But we need to very heavily pushed them towards adding people to their trust circle. And again, once they add people, you want curators that have large breadth, right? They've made a lot of comments or claims about a lot of topics or, excuse me, a lot of atoms in that topic because really you want the user's feed to be littered with insights. Um, so yeah, trying to think. Yeah. Oh, I should mention this is important. Uh, so a big question. You'll notice a lot of what I've done here. You can, you're making decisions and you're staking, you're making claims. I haven't added a lot of functionality for, well, I'm not encouraging the user to put heavier or lighter claims based on conviction yet. I don't know if you guys are doing that a lot yet, but so it's almost like the quote unquote. Price discovery almost for trust tokens in my app has not occurred yet. So a big question and actually I want I would love feedback from you guys on this one. Um, at what point should I be pushing users more like, hey, steak more for more conviction, right? So because you have the curators wait for their claim, you also have the user's weight on how much they trust a person, their trust circle, I guess. You know, my knee jerk reaction is, okay. That must mean that eventually, once we get this going, we'll probably multiply. You know, their their trust for the person times the, uh, the weight of that trust, trusted person's claim to kind of come up with, I don't know, some sort of value. Maybe they'll tell will order. I would think that's that's a big part of it when you can only show a few icons in the feed, for example. Um, obviously the ones that have the most weight should probably be towards the top. So right now, there's not much of a push for users to stake a lot. Um, because I'm not using the stake weight that much, if that makes sense. Um, that's a process that will be happening. I think once we're more confident of which way it's heading. But what are your guys thoughts? Um, is this a point where we should start? How would we be using the weights? Would we use the way that I mentioned it? Or what other ideas do you guys have? I think there's a good question for Billy too, if he's listening. I think you could you could do it progressively like I would. The first way I would introduce it would be using the weights as a way for people to kind of staggering people against each other. So it's like, I want to I want to trust Kylan more than this other person. I want to trust both, both of them. But I want to trust Kylan more. And the way that I would stack rank them is by staking like a little bit more. So I think that's like very small incremental step in the direction and then getting how do you expect that to be? What, what if, what effect do you expect that to have on the user experience? So if I stake more on you over some random person, but you're both in my trust circle, for example, what effect would do you think that should have? It depends on the context in which you want to consume it. Which like brings me to it. Like I, I really think that locking down the ontology is very important so that these are like composable across applications, but I would think that, you know, um, the, I can think of it in your, I'll think more about it in your context, but like, what immediately comes to mind for me is, you know, something showing up in a feed or if I need, like, it's like, okay, if someone's like higher my trust circle, maybe I see more of their things than someone who's lower in that trust circle. Or if you want to, you know, render, let's say you've got like three slots for profile pictures on the front end where it's like, okay, which people from my trust circle trust this thing? Yeah. It's like, okay, well, I want to see the picture of the guy who's like top, like the top person. Like, so just, just things like that where it's like, if you got to make a decision of like what to surface to the user, I think the stack ranking of how much you trust someone in a specific context, like that's when it starts coming into play. Yeah. Yeah. There's a lot, But I guess when you say it like that, there's there's a lot of ways you can do it. Like it's just this, there's always been this tricky thing in the back of my mind because whenever we try to make some sort of ranking, when it's a public facing ranking, that's when it gets it seems like that's kind of when it gets screwed up, right? If we want to say, oh, which curator has the most stake, are some dudes going to go make another account and put, you know, one hundred thousand trust on himself and say, look at me, I'm the top curator. Um, but so maybe it is more just in the personalized context when it would actually make the most sense. Yeah. I mean, I think like, yes, but also I think. I'm, I'm a big fan of. Kind of the. Keynesian beauty contest problem and like. The thought that if there is economic weight behind it and you kind of surface the ability for people to speculate on something and the game is, you know, like there's, there's like, I want to. I want to take an economic bet on who I think everyone else is the most trustworthy person in this context, rather than who I personally think. And through that dynamic, you actually surface better signal where it's like, it doesn't matter who I trust, it matters who I think everyone else trusts. And that ultimately leads to better results than people subjectively choosing who they personally trust. So it's hard to get to that point because, you know, like God, we're still a small ecosystem. People are going to gain whatever you have and like it, it kind of like requires, you know, I think the higher the volume, the, the more people participating, the more that kind of plays out. So I don't know how it plays out in a super small concentrated setting like we have. But the counterpoint to that is also, I think the people interacting with it with things right now, like hopefully have those things best interest in mind. So maybe it gets gamed less than it otherwise. And I'm like super adversarial, larger environment. Yeah. Yeah. I've thought about if, if, if they are, if the statistics, let's say, oh, who has the most followers as a curator? Um, I think as long as like, it's just tricky because sometimes you want to put up, oh, who has the most followers? But yeah, people will game it, but there are, I guess there are sometimes opportunities for sometimes it's not all noise. Yeah. One, one one thing to think about is we're working. It's not going to be ready soon, but soon is relative. But like, we're working on allowing people to claim Adam wallets through the Adam Warden, um, like very rudimentary versions of that already exist, but like we're building out more advanced kind of claiming slash recovery functionality. So the reason I'm bringing this up is because, okay, imagine, um, like the goal is going to be adoption. It's going to be growth, it's going to be distribution. So it's like, okay, how do we, how do you get kind of like relevant people interacting with what you've built? And one way is through, I think through like trust circles. So imagine lots of people have lots of trust volume being staked on the fact that this person is trustworthy. So imagine it's Vitalik and everyone's claiming that Vitalik is really reputable in the context of Web3. Um, like the atom wallet representing Vitalik is going to be accumulating fees. Uh, and maybe Vitalik doesn't care so much about money, so maybe, maybe it's some other guy, but like, let's say Vitalik cares about money then. And Vitalik, the atom wallet representing Vitalik has accumulated a ton in fees through people saying that he's trustworthy in the context of Ethereum. Um, now you can reach out to him and be like, hey, like, why don't you onboarded this app? Because, you know, people are claiming stuff about you and if you come in and claim your thing, then you claim it at a mall, then you claim all this money and all the fees that are being generated through people making claims about you. Yeah. So I'm trying to think here. So it would be the atom for, for example, his ENS name or his crypto address. Right? Yeah. What we're also working on, like, uh, James, are you still are you there? Do you know, like which methods were we're introducing for Adam wallet recovery at the beginning? I know we're doing like. Uh, yeah. So we're, we're starting out with, with DNS, so you'll be able to do that, um, and add like a text record to your domain and have claim ownership over an Adam wallet that points to that same URL. Uh, this one will be a little bit longer, but we're working out ways to log in with OAuth. So to prove that you either own like a Twitter account or you own a Spotify song, we're going to have this off chain service that authenticates, and then we'll give users basically a signed message that then allows them to claim a specific wallet. But that back end will be flexible. Uh, it doesn't matter like what you verify. We just always give out the same signature so new things can be added over time. Um, so you got, you know, as a community, you want to fork this or, I don't know, it's a little different because we're, we're the owners of the, the atom wallet, but we can figure out a way to maybe collaborate and kind of scale this out. So real quick one there. Thank you. Um, so my, my one concern, my immediate thought when I think of people claiming for their atom is you're, this is a potential PR issue, right? If someone makes a controversial, if they're controversial figure that everyone talks about, does that automatically think mean they should make money off of it? Right. If I'm, you know, some, I don't know, bad person, bad media influencer or a, or a, a bad politician, some evil doer out there who's infamous and, and everyone's talking about me. Then then should I be able to make money? Right? Because the, the, the news headlines are going to say intuition. Blockchain is profitable for bad people. That's what it's going to say. You know, do we have a mitigation against that? I mean, I don't think there's a there's there's no way around it in something like this. And I mean, if you look at life generally, it's like, aw, let's take Twitter as a less extreme example. It's like, okay, Twitter's not looking at how bad you are when they pay out their monthly fees. It's just like how many, how many impressions did you get? Like how many eyeballs saw your thing? And they say, you know, no press is bad press. It's like, how are you going to how who determines who's good or bad or like that? The community hates it. It's like if you do that, you enter this world of censorship. So I think it's just it's just relevancy. So if you're relevant enough to the point where people are like claiming you're a bad person, at least you're relevant. And that's interesting. It's, it's interesting enough to these people to put money on it. Like that's valuable to the system. So. I don't think I don't think we like. Yeah, I don't think we observe that as like. I think now that I think about it. So there's, there's definitely some mitigations like. I think if people know that engagement is profitable, then I wouldn't be surprised if you start seeing convention where people kind of try to keep their staking amounts on bad atoms lower. Of course, that means that someone can come and influence it more easily. Um, but the other thing is, yeah, I guess you're right. X does pay money on engagement. As long as our network is a net positive, then we can make the argument that whatever good atom amount, excuse me, whatever good atom revenue someone gets, uh, should outweigh the negative. So I guess there's an argument there as well. All right, let's see. All right. This one says when you talk about curator overlap, two curators of testing on the same atom. I was thinking the overlap could actually show right there in the queuing step, like a little trust score badge on each claim before the user submits, powered by the trust. So instead of just seeing an atom, you see. Okay, so you'll see if you make a claim, see something like ah, you know, your buddy also made a claim. Yeah. That's good. It'll tell you kind of who you agree with, who you don't agree with. Um, and generally we also want once users start making their own claims, then we can start matching them with curators who think a lot like them. The way that I see it is if you pick the right trust circle. So theoretically, let's say I make a bunch of claims about crypto, and I'm not a curator. And then I see a curator who happens. You know, I've made let's say I've made a thousand claims on crypto, and then I. My, uh, hive mind says, oh, look at this curator. Uh oh, Billy, you and you and Billy. Agree on all these overlapping claims one hundred, one thousand out of one thousand times. It's very powerful for me at. That point to add Billy to my trust circle, because as long as I have a bunch of people who agree with me have a lot of affinity, uh, with my opinions, then essentially your trust circle kind of becomes a kind of almost precognition instead of you having to figure out what an X account is or who they are, if they're trustworthy. Essentially, your trust circle is you, right? It's a very close model of your opinions. So I think that's really powerful. I think it's that that would provide a lot of information to. You very quickly. So I do think the trust circle functionality that intuition is going for. In general, it seems like people are heading in that direction. I think it's excellent. Um, one thing you know, another thing I should add is that my I have a little concern when I'm doing the curators when I'm offering curators, you know, I was thinking of doing a starter pack, but that means that hive mind, when we encourage people to make their trust circle, you actually don't want one DApp being too influential on a trust circle. So you also want other DApps promoting curators for the same or similar topics so that at the end of the day, a user's trust circle is kind of an aggregate of who they trust on topics from within different contexts within hive mind, right? Which is probably going to start off mostly talking about Twitter accounts. Um, or if, you know, maybe they're on another app that works on agent ranking, right? Um, so, and of course, obviously there's crossover, you trust someone for their opinions on, on agent ranking, then you're into rank. Then if you come to hive mind, you'll have those trust circles, which is good. I think you want larger trust circles because as long as you actually do trust those people, you'll have a lot, uh, larger coverage. Um, and obviously larger coverage also means the curator themselves. Uh, in fact, let me show you one thing I got. I got one more thing to show, I guess, uh, tire screen. So if I actually go here and I am smiling, Kylan and this is not done, the explorer stuff in general is just a bit messy. There's just a lot of stuff. It's not really super organized. I haven't really had my designer work on that too much yet. Uh, so if I go to my profile, uh, it doesn't show all of it. Um, yeah, let me see here. So it does show you kind of your impact. Okay. I have crypto here of six followers. Um, let me see if I can get this. I have too many wallets, so I don't remember which one. Try this one, I guess for my profile. Uh, anyway, uh, we do have some widgets that kind of tell you, hey, you know, you're a curator for this topic. Do you want to like you need, you need more claims, essentially, like you're, you're, you're not quite there yet. Um, all right. Well, it's not on here, but I thought it was, uh, but yeah, so that's the good thing is I was worried that that hive mind would become too. It would centralize people's trust circles too much. But as long as other daps are also pushing trust circles again with the same triples and conventions, then, uh, that should have a decentralizing effect, which is exactly what we want. Now let's see here. Okay. That's good. Okay. Got any other major questions here? Yeah, I think like in terms of like trust circle semantics, um, can you pull up again what you're using? Yeah, yeah, yeah, let's do that. This is a great time to do that. Um, so I'm just using this address trusted for topic. Um, yes, you could do, what do you call it? Compound or, uh. yeah, a triple within a triple. But, um, I was trying to keep it simple for now. Um, so I'm hoping this sort of structure makes sense. Is there. I know, like, does anyone. I'm just curious if anyone else has like perspectives here also. Well, one, I'm curious, like, what is the Adam trusted for look like? Is it just the string trusted for or is it something else? Um, I guess let's pause there. Yeah. So I pretty much always try to do a quote unquote text object when I can. I don't, I don't like doing the IPFs ones if I don't have to. Uh, so if you literally do just trusted for the, just the basic string, that's the atom that I'm using as far as I'm concerned. I try to go for the simplest atoms when it makes sense. I don't know if that's a good strategy overall, but I think so far it has been. Okay, so how does that compare to like the predicate registry standard that we're. Like, I don't know, are they published? Slash if they are or if even if not like. Is this a convention we want folks to follow? Is this okay? Can we like make. If people start using. Just like the text object versions is. Can we make them compatible? I'm just curious. Yeah. So, I mean, ideally we kind of move towards what we published. It is published, it's available. And I can explain the reasons why. Like so we kind of work through, you know, what are the pros and cons of strings versus an object? And if we store an object, you know, we are landing on is the schema.org defined term. And then what that allows back end systems to do who are indexing. That is, to know that this string is is a special predicate, that it has some enshrined values that like you should treat it this way, you can index it better. Um, if it's just a string, it's kind of like hard to pick up like, oh, this is a special predicate unless you already know it ahead of time. And you know that constant, but say there's like fifty plus predicates and they're all using this defined term part. That means anybody who's indexing our backend back end knows that like this predicate does something. It's used somewhere to, to, you know, change a front end, change a back end search, something like that. Um, and so obviously we do have strings now that are kind of legacy. There's the hash tag, it's the one that you're using. And in small cases, we can kind of have legacy support, but ideally we all move towards this, you know, object. And to your point about IPFs, it doesn't necessarily have to be on IPFs. We can support strings obviously just in like Json format on the atom data. Um, that obviously gives you a different atom ID. So we're going to have to pick one or the other kind of long term, but that's the thinking of, of why move towards this object as opposed to just a raw string? Yes, yes. So, uh. Thank you. Um, yeah. So with, with, I guess if you're not doing it to IPFs, then that does change it. I guess my question immediately would just be, let's say I insert a Json string into the data field or URI, whatever you want to call it. And then what? I'm just trying to think, how quickly does that get indexed? Does that get indexed any slower or faster than a regular simple text? My, my issue with the IPFs one specifically is that they kind of have to propagate. And from a user experience perspective, if, if I need. I need prompt responsiveness. Someone makes a claim or ask someone to their trust circle whatever I want it to like immediately start taking effect. It has to go to the blockchain. That's fine. Um, but if we have to start going to IPFs and we don't know how big the file is, then all of a sudden that it just gets way too slow. Um, but so I guess my question would be if you're doing the Json stringified version, is that a lot faster? It should be the same speed as a regular string, right? Because you're, you're just parsing it. You don't have to do any, any lookup. You basically just have one other function, which is like, turn this string into a Json object and check it. You're talking like milliseconds there. Um, but obviously there's like a character limit of one thousand right now. We can bump that on the smart contracts. So this works really well with predicate objects where they're most likely going to be below, you know, a thousand characters. Um, it starts to not work if the atom is very big, say like we ran across this problem with Twitter. So like then you actually have to post it to IPFs and then use that URI. But for predicates, I feel like there's, you probably don't need IPFs and most of those I can, I can run the math and see if they're under our character limit. But most of the time that like, yeah, if you just use a string, you get the benefits of not having to have that, that long lag of. Yeah. Then resolving with IPFs and some gateways, they take like, you know, forty or 50s, it can be an extremely long wait to get that that file resolved sometimes. Can you tell, you tell me more about this Reg, more about more about this because I don't know much about it. I haven't heard much about it. Who's in charge of it? How far along is it? Where are you guys discussing it? Because it would be nice if we had kind of a very basic, uh, hey, these are kind of the atoms that most DApp developers think that we should use. Um, I don't know if we'd use intuition itself for that. I think that's an interesting question. Or if this is just how is this conversation being had I. And is it. Yeah. So we just kind of published stuff last week. There is an intuition data structures folder that has like basically the just markdown explanation of what's going on. Uh, we did codify that last week. We haven't published those packages, but we're getting ready to do that like this week. Um, and so what you'll be able to do is import, uh, intuition slash predicates. And then it has like these top one hundred that we're recommending. Um, and it'll, you can easily import them, get the ID, get the original data. And in the future, we're even going to do multilingual support. So we start to have these, these packages that you can just import and kind of see what these predicates will, will be used for. But right now it's in the link that, that, that put in the chat. Uh, but hopefully in the next week or two, we will actually have npm packages that you can import and easily use these in your, in your app. Awesome. Get off the ontology up. I can showcase that quickly. But yeah, that's that's an initiative. Like to actually do that to just explain this web page ontology dot intuition box is like you leveraging, um, leveraging intuition to have like a registry of the used predicate for different use cases. So right now, like it's just like a, a proof of concept, like highlighting like web two, um, registry. But yeah, next step would be like to bring it like a, Connect it to to intuition because obviously, like most of the job, like, uh, need to comes like, uh, from the team and the job from, uh, Kim's putting like the list of, uh, predicates like is needed, but then like this need to evolve also because like, we can't like cover that, uh, from like the one single standard to cover all use case. So something need to evolve and be flexible for every project to say what they want to recommend. Yeah. Let me mention real quick. So this actually reminds me, I actually think it makes sense for us to create an intuition topic itself. And, um, because it's a crypto itself is a very broad topic. We actually need one, I think for intuition where, so it would be interesting, like, um, you could use trusted circles where the topic is intuition for. Hey, what's the right atoms to use or what? You know, whatever opinions you want. Even like the smart contract. Um, if you guys have seen the MetaMask snap, if we it would be good for us to start tagging the smart contracts that we're using, for example. But that means that we would probably want to use a trusted circle for it. Um, so yeah, that means we'll have to create an intuition topic, which would be kind of cool because we could probably almost all be curators of that one, right? So that actually might be a lot of fun to, to create that early on. Um, start eating our dog food, as they say. Okay, cool. Well, that's it for me. Um, I posted a link in the chat if you go to hivemind HQ, I. Oh, oh, we're on my. Let's see here I'm on HQ, I oh, you can get more info. Um, it doesn't look very pretty, but if you go to the curators section, if you want early access. I know a bunch of you guys have signed up. As you can tell, I'm still ironing out the kind of the final details. Uh, so I do want to get you guys in there if we can start using it soon. And those early access people. Yeah, they would have access to it. And, um, it'll, it'll, it'll still have bugs. It'll be uglier. Uh, but I think we want to get the network graph kind of going and get early curators in there populating the network graph so that once we eventually do a larger launch, once we get to that point, then, um, then we'll already have a bunch of the network graph populated. Me. All right, that's it for me for the demo. If you have more questions, let me know. I am going to be posting in the Atlas Discourse message board thing. Uh, once we have this video, I'll probably put a link to it there, maybe even embed it. And I think I'll probably mention a bunch of the topics we had here and some of the outstanding questions so that we can continue the conversation. Um, yeah, thanks for coming out. Um, that demo went more smooth than I was expecting, which is nice. And, uh, thanks for, uh, tolerating the, the different time zone. I know, I know it's very late for a bunch of you guys. It's just, I'm in Japan and there's, it's a terrible time zone. Uh, so thank you for for tolerating that. Yeah. You know, it's kind of hard, um, being a worldwide community, right? It's always hard to get everyone together. So huge, huge shout out to everyone who did come, especially if it's late or early, no matter where you are. Um, Kylie, there was one more thing that I wanted to surface that Luda had said early on in the chat where he said, um, when you talked about curator overlap, two curators attesting on the same atom, he was thinking that overlap could actually show up right there in the queuing system. So like a little trust score badge on each claim before the user submits, powered by the trust MCP. So instead of just seeing the atom, you see how strongly your trust circle has already backed it. And does that make sense? Yeah, I think that makes sense. And so one thing we've been thinking about, if I go real quick, we definitely want the user's opinion on whatever topics because that's actually how we help make sure that they actually agree with the curators that they have. So if I go in here this way in is another way that we can do it, where we can say, hey, we can start presenting claims here that a bunch of their trust circle people for a specific topic have made claims on that signals to us that that claim is related to that topic. And then we can, this user follows people for that topic. So we can say, hey, do you agree or disagree with it? Um. I guess that's not the same thing as what he was saying though. Um, yeah, there's also social proof when you're making claims. Yeah, you could have. Okay, let's see here. See, I make a claim so it could be like, oh, these people use trust circle also agree. I just don't know if that's like, sometimes I wonder how much how important it is. I haven't got the answer to how important it is for to know a user's opinion on a claim once their trust circle has already made that claim. I mean, it's good to know if they disagree, but if they agree, how does that that doesn't change much in their feed. Like, yeah, it'll show the trust avatars of their trusted circle people. Um, and that's probably enough. Um, like, do you need to be reminded that you think, uh, that you think an account is funny or entertaining? You might already know that. So to me, there's a question of like, how useful is it to make to, to pile on to claims that other people already your trust circle already? Well, actually, I take it back. I take it back because once you have the profit motive with the bonding curves, then you can have a feed that just says, hey, look at, uh, your trust circle, uh, said this about this, you know, person today. And that could be if you think that that more people are going to stake because this curator is ahead of the curve. Um, then I guess you would want to stay because it could be profitable. Yeah. There's so much mechanics with this stuff. It's I love it. It's very rich. Lots of data that we can work with. Uh, but it does make predicting it, uh, pretty much impossible, but it does make it exciting, right? Yeah, definitely makes it exciting. Kind of what you're talking about. Yeah. Yeah. So we'll see. Um, hopefully that's a good follow up. If not. Of course you guys can always chat more. I know it's probably more async than it is on a, on a, on a sync call together, but ways to keep in touch. Um, this is actually really cool. Thank you, Kylan for, for taking the time to demo hive mind to us. It's actually something that myself and Billy had just chatted about briefly last week about wanting to start seeing some demos from everybody. So I'd love for us to continue these calls and maybe we can schedule for like each person to give a good demo of what you're building. Um, every Monday and we can kind of focus on one product and then also leave some time there for, for blockers or for, um, uh, convention chatting. You know, if anyone else. Yeah, if anyone else wanted to give a demo today, I'm, I'm still here. I'm sure some more people would be here, but I also know everyone if you want a break, I get that too. Um, so just kind of leave the floor open or if you had anything that you specifically wanted to chat about, anything you wanted to bring up, questions that you have, anything you're concerned about. And I'd love to open the floor to Z or Salo if they have any thoughts from the from the Builders Guild side of things. No, thanks for the demo. I think like from my side, like I really like the the way you are approaching it. Going with the circle and maybe like in the first term, like not trying to compute like a, a score ranking, just like filtering by who you trust. And I feel like that's really how you, you should start. So yeah. Love it. Yeah, I agree with that. Nothing to add. But I second these words. Yeah. Hey, man, it would be great if we could just get a global trust score. And you should look at a glance. Can I trust this person or not? But that's unfortunately not the way the world really works, right? I saw a funny comment. Um, it was I don't remember the saying, but it was like as soon as, uh. What is it? As soon as a signal starts to get measured, like it starts to get manipulated or it was something like once the, once the number becomes a metric, then then it's no longer reliable, essentially. Yeah. It's, it's, it's, it's a Schrodinger's Schrodinger's cat sort of thing or something, I don't know. I, I agree with that. And I also disagree. I, I think like the, like a global score, universal one is the wrong way because for the, the reason you mentioned, but also I think like having like the trust circle as a basis, we can do awesome things because like with your experience, like someone entering your application will be limited to what is trust queries. But then like, if it doesn't have like a trust circle, then like baby, like just following one or two people from intuition who already have like, um, their own trust circle. That's where you can like derive recommendation from not people you trust directly, but indirectly. And then if you want to scale that, instead of having like just recommendation as you have, you can like decline a trust score on like indirect relationship with someone. That's what Luda is, uh, is working on with, uh, with his MCP and trust score computation. So I feel like there is a way to, uh, there is a lot of way to, uh, to expand that after. Yeah. And I think it's interesting. So my initial thoughts too are. So I thought about adding. People, your trust circle trust, right? So one, one extra skip. Um, obviously the queries can get a lot more difficult. And there was a little, I was a little bit worried too. Okay, what if someone just adds a bunch of garbage trust circle people? But the truth is, if, if curators start to act up, uh, people can just unfollow them and that will mean that they'll lose influence. I think the curator position will probably be somewhat lucrative. So I don't think people will want to screw that up. Uh, so I think making sure that users know if any curators doing something that makes the user experience worse, like making low quality claims or false claims, or trusting people in their own trust circle that don't, that aren't, that aren't good, essentially, um, making sure that users know which curators are kind of, quote unquote bad for their feed and letting them remove them. So that's another thing. So if you start, if, if you as a user, have a curator where you and that curator start disagreeing a lot, then Will probably notify the user. I don't know exactly what that'll that signals, but that's something we'll find out as we actually start using it more. So. Yeah. But the, the whole, the, the extra skip with the calculations. That's why I'm going for, I want the user to have a lot of their own curators rather than having to go the extra hop. That's how I decided to tackle it. And that's just for me personally. That's. I think it works better for me. Yeah. There was one other question that was in the chat, um, uh, just shortly after you got started from Passive Records or Sam from the Sofia team, he said, given the Trust circle feature, would you describe your app hivemind as a social app? Yeah. So around the time when I, I don't know if you guys remember, but for my hive mind discord server, I was playing around with this bot and, um, I realized, so you could ask the bot to give you like, you know, that's kind of a summary of the claims on an account. And so I put that up there and I, I quickly noticed, uh, that first of all, the things that you guys, I'm going to say, you guys, not you guys specifically, but generally what people were looking up were themselves. They're either looking up. Oh, what's the reputation? What's my reputation? Right? That's the most interesting thing. Then also looking up kind of their friends. And I think at that point, that point, I started to realize that there is a lot more of a social aspect to intuition. And I, I actually think that's probably the best way to, to, to present intuition to people because social media is just way more engaging. If I look at the apps and the websites I spend the most time on during the course of the day, it's going to be YouTube, Twitter, and Reddit. Um, so really that's kind of what people gravitate towards. And if you're making claims about stuff that people either don't care that much about, or it's not that compelling, or it's not what's in front of them. It's not the topics that are in front of them during the course of the day. Then they don't have they might not have much of a reason to go to your app. So that's why I've been using the browser extension is I say, hey, you're already in Twitter, I just have this browser extension installed. It'll be somewhat passive. Um, and people are always on social media. So that's definitely the area that I decided to target. Um, and part of that's just because, um, that's just what's compelling to users. Uh, you know, it's, it's a cynical sort of, uh, thing to do. And in fact, whatever, whatever social parts I can add, I will tend to lean heavily into adding them. Um, a lot of the patterns we're using, like, oh, this person also likes this or this. You know, matching people through kind of inference and connecting the dots. A lot of the stuff that I'm doing, you see similar stuff in like Facebook, right? Instead of curators, they're friends, right? We can recommend a curator of a curator or a trusted person of a trusted person. That's the same way. Facebook says, hey, you may also know this person because they're friends with three of your other friends, right? So I'm borrowing as much patterns from that as I can because social media clearly is effective. So I'm not trying to fight that. Yeah. The more social the app is, probably the more engagement we'll have. Agreed with you there on on just about every front being social, making things fun, making things connect to others. Yeah. Very huge. What if I say there really isn't a global trust score? The trust score should be based on the end user settings. This is where staking and waiting really shines in my opinion. How much you trust someone or something changes dramatically when you change one or two variables. Yeah. You know, while we're on the. Oh, sorry. No. You're good. Yeah, yeah. There's not much of a trust score. It'd be nice if we would have had the Holy Grail, but that was one of the biggest questions people would say though. They would immediately be like, oh, we can't just anyone just make an account or can't I just. Yeah, there's a lot of ways you can kind of gain things. The truth is, behind the scenes, the DApps, a lot of our projects are probably going to do very heavy calculations behind the scenes and not let everyone know that we're doing that, because we'll be able to find a lot of insights. I think when you tell the users, hey, don't worry about the global scores, just focus on your trust circle. I think organically you'll get some strong signals if you crunch the numbers that way. But the second you start telling them that you care about the global numbers, then they start manipulating them. So it's kind of this, hey, don't worry about the global numbers. But also I'm probably going to look at the global numbers, right? Yeah, yeah. Always things are on the back end. And Thibaw had said a good idea for the browser extensions intuition to create applications that act as additional layers to this trust score and encourage users to participate in all areas, whether to express support or disagreement in order to gain a genuine perspective. Or as needed. Yeah. Vital said that as far as like the trust score stuff, right? It's in all or it's all in the query parameters like you were saying? Yeah. You know what? While I have you guys, another important question is this kind of I'm thinking about what you're talking about. Oh, it's social app. So that's a it's a good thing. It's compelling. The, the one tricky thing is I'm noticing, I don't know exactly what form the. I'm going to share my screen again. What form disagreement is going to look like, right? So if I go because there's three kind of ways that I've thought that disagreement might look like. So if I let's see here. Okay. So disagreement. So we have the first one is the most obvious of all versus counter vault. Uh, second one would be negative claims. And then the last one would be omission. So what I mean by this is, okay, Billy is trustworthy. So obviously there's you can do for or against that. Um, the one thing is right now most things are for. Right? So on something like this where it's like a person claiming that. Billy. How do I say this? Staking against the claim that Billy is trustworthy. That's an inflammatory that's a fairly inflammatory claim, right? So our curators, especially curators, because curators are, are going to become well enough known that eventually the person saying that Billy's not trustworthy. Billy's going to find out about that. In fact, he's probably going to know that very quickly. And he's going to ask himself, who is this person? So doing that is very confrontational. So I think you're going to see some resistance, uh, to claims that are related to people or that are not not highly contested ones. Right. If I said Craig Wright as Satoshi, you, you know, you'll get a bunch of people disagreeing and some people agreeing. That's what you expect. It's a contested claim. Um, the other thing is negative claims. Instead of instead of making the claim Billy is trustworthy. Someone might just straight up do. Billy is a scammer. Um, that's theoretically even more inflammatory, right? So I think you're probably not going to see too many people do that, but what, what I would expect you to see and I would, or at least I would be surprised if it's like this is let's say most founders for crypto projects have trustworthy, um, builder. You know, these are just tags, right? So if you don't see much stake on on Billy is trustworthy. Like if it's not towards the top. And for most founders, then these usually are the top. Then it's almost like the omission, the fact that it isn't there is kind of the signal, if that makes sense. It's a lot more subtle, but it's less confrontational. Like if I go to a founder and his top claims are, oh, he's funny, he's entertaining and he's, um, you know, he's, uh, very fashionable. That is the red flag, right? Because they're not related to building. So I think that was, you know, we don't have to go deep into this topic, but that's a topic I've been starting to think about a little bit more is like, well, what does disagreement even look like? Because we need to make sure that people can express disagreement in some way on things that are important, like the reputation of a, of a, of a crypto founder. We want to make sure they can do that without feeling like there's going to be retaliation. Yeah, it's definitely interesting to think about. Uh, did you see this message that just dropped into chat from Believe and Tylenol together? Yeah. Okay. Do we have multiple. Should we do first or which one? Yeah. Let's go with Luda. Um, so have you ever thought about surfacing a real time trust rating on X accounts directly in the extension? So similar to what ethos does, you know, ethos has these reputation scores for people on X, but instead powered by your trust circle instead of a global score. So it's still a score. Uh, yeah. Uh, let me rephrase that. So earlier, kind of what I was trying to think or to do was like, okay. Your trust. Wait for a curator multiplied by the, uh, the weight of their claim. So if they if they trust Billy and I trust them, then we multiply the weight of the different stakes. So I think that could be the closest thing to a score. Um, but it also, you know, which claim I assume it would be hashtag trustworthy. We're talking trust score. I'm not against it. Um, a part of me worries a little bit. Sometimes when you start leaning heavy into that, then it becomes some sort of, yeah, everyone wants to game it or like it becomes a little bit. There's a lot of vanity, I guess, and that makes me a little bit concerned. Um, I know that's like Instagram, I think started limiting the, I think they didn't show like numbers. How many likes on a post. They started hiding some of that because they felt it just became a big contest. And it was if it becomes unhealthy for the system, then at that point you can revert it. But yeah, I'd be very open to doing that. Um, I guess my question is, do you guys find those scores are helpful? And we'll give you a chance to talk here, but I'll kind of fill in from my own point of view is like, I think they're helpful sometimes in some ways. So for from my own perspective, right, like when I scroll X, I use, I've used a number of different apps over time, but currently I believe what I'm using is the wall chain. Uh, yeah, the wall chain X score. And so I'll share my screen real quick. That way you have that way you can actually see what I see. Um. But if I'm here on X, I see This green score next to people's names, and it's more or less a reputation score based on it's based on who is following you. So like I have this score of one hundred and thirty based on the people who follow me. And that's based on the people who follow them. And I'm, you know, there's some back end analytics there as well. And if we look at like intuition has one hundred and eighty nine. Elon Musk has a score of one thousand, which is the overall top score. If we look at Billy right. Billy is sitting at three hundred and twenty six. So these are things like these are signals to me. Yeah. It's just the trust predicate. Essentially. Yeah. I mean, if it were to be built on intuition, it would be something like it would be somehow ranked on people saying that they trust somebody. Um, and so like the way that this is calculated from wall chain is I don't know all of the specifics, but it's like Billy is followed by a lot of very influential people who are considered trustworthy, thanks due in part to the people who already follow them or the things that they've done in the space. I don't know all of the backend details of how these were calculated, but I can say at this point, right? Like as I continue to get more and more followers who are considered trustworthy or big accounts, um, or reputable accounts. My score here goes up over time. Ethos does something that's similar. Yeah. Uh, real quick. So this is built on intuition or not built on intuition. You said no or something. Okay. Yeah. It's called well, chain is a browser extension or. Yeah, it's a browser extension. Um Right here. Well, I'm very open. I'm very open to adding this sort of functionality if people feel like they want it there. Yeah. The one when I think about scores too, um, I've been thinking more about scores as like a, I think I like the idea of having as a percentile because I'll be honest, you, you tell me one thirty was I have no idea what on earth one thirty means. I have no idea what twenty five means. Um, like and then one's a thousand. So is that like how many times better is that? You know what I mean? Than one hundred and thirty. Is it seven times better? What does that even mean? So yeah, I think percentiles are probably the way to go when you have to do some sort of almost global score. Um, because then at least you can kind of compare it like, oh, I'm in the top whatever percentile. But yeah, I'm open to doing stuff like this. Yeah, yeah. Cool. And that's what I wanted. Oh, yeah. Yeah. I was gonna say, this is why I want it. This is why I want this to start getting into people's hands, because then as people use it, they can start giving me more ideas because and that's why I try to do this demo as quickly as I could, because it's just way better to, um, start getting feedback. I can't guarantee that I'll make stuff that users want, right? So I need user feedback, um, to make sure I'm actually heading in the right direction. So yeah, sure. And, um, you know, Luda had followed up here saying that he thinks context matters a lot, that a global score feels arbitrary. But if it's coming, if it's a score that's based on and coming from your own trust circle, it feels a lot more personal and relevant. So like it's telling you what the people you already trust think of this account, not just based on an anonymous crowd, pretty much. If you want the score to be so that people can't like, screenshot it essentially, because the second people screenshot and say, look how look how good my scores. And everyone just wants a high score and nothing more. But when it's different for every user, the score shows different based on who's logged in. Then it's right. It takes away that which is good, which is good. It's more personalized. That's just a good thing, you know? Yeah. No, absolutely I agree. Okay. So believe is saying, uh, he wants to know if hive mind can do predictive analysis on claims slash identities for easy decision making for users. So like if they see a claim, it sounds like if they see a claim, would we tell them, hey, what? Since you have trust circle who staked a certain way on a claim that maybe you should stake that way? I think he's what he's trying to say. I think that's what I'm getting. Yeah. Um. Yeah, there's a bunch of things you can do. And I think it I don't even know if you need much analysis other than, hey, most of your trust circle says they went this way. So you you probably see. But I don't know if we want to encourage them to go one way or the other, because if they disagree with their trust circle, that's a strong signal that we want to know. Um, so the question becomes, will that influence the way that they stake to, right? If you're like, hey, these, these people stake this way because you do want people to give their honest opinions as well. Lots of food for thought. Yeah, we're going on an hour forty so we can start wrapping up if you want. Yeah. I mean, that's what I'm kind of thinking, you know, and I know a lot of our European friends are starting to head offline. Uh, I know we missed some of them in, in hindsight, if they watched this. Thank you all for being here. A lot of the a couple of the core team members have left as well. But I'm very, very appreciative for James and Joji coming by. Billy and Vidal are still here. Thank you all for your time very, very much and your thoughts and opinions. Um, I guess last last couple minutes here, anyone have any last last minute thoughts they want to call out throw out there before we wrap up and then we can, we'll start coming back to this on a weekly basis, getting everyone together, being able to chat about these things. Yeah. Don't forget hivemind, HQ dot io curators. That's the best way to get early access. And then also in discord, I have a hive mind server. So that's one place where we're going to start. That's probably where we're going to forward people to discuss, uh, the latest with the extension. We also have the explore the bot, the snap. Um, so yeah, if you want to join that server? I'll include it. Oh, if you just go to my website and then down here, discord, you can also go to our X account as well. If you want to follow. So thank you, I appreciate it. I think this demo went really well. I'm very happy with how it went. And um, I think we went over a lot of important topics, so I'm stoked. And woods, thanks for putting this together and yeah, thanks everyone for coming out. Yeah. No, thank you so much. It was awesome. Very cool to see Hive Mind from your perspective and how things are going so far. Hear a bit from you and your thought process around it. Thank you everyone for coming out. Awesome time spent here. Thank you very much for coming, for hanging out with us. Everybody else really appreciate you guys believe. Thanks for dropping some questions. Would love to build on intuition through hive mind if that becomes possible in the future? That's pretty cool. Yeah. Hey, listen, if we get this functionality working well for X-Com, then that means we will want to add it to a bunch of other places. Like I said, YouTube or I don't know, IMDb or LinkedIn or wherever. Um, we want this to be able to be applied to, to everywhere. Really. Right. So, um, yeah, as long as things keep growing, it's, uh, the builder team at Hive Mind will grow. Right now, it's pretty much just me. Yeah. No, I get it. Well, cool. Thank you. Thank you for your time again, Kylen. Thanks, guys. Everybody here. Vital fungi. Bill. Jenny. Appreciate y'all. We will see you around. Reach out if you need anything. And, uh, we'll see you next Monday. Do you trust everybody?