Alex Gleason was one of the main architects behind Donald Trump's Truth Social. Now he focuses on the intersection of nostr, ai, and bitcoin. We dive deep into how he thinks about the future of nostr and vibe coding: using ai tools to rapidly prototype and ship apps with simple text based prompts.
Alex on Nostr: https://primal.net/p/nprofile1qqsqgc0uhmxycvm5gwvn944c7yfxnnxm0nyh8tt62zhrvtd3xkj8fhggpt7fy
Stacks: https://getstacks.dev/
EPISODE: 164
BLOCK: 901101
PRICE: 957 sats per dollar
(00:00:02) Alex's Presentation at the Oslo Freedom Forum
(00:01:31) Challenges and Opportunities in Decentralized Platforms
(00:02:31) The Role of AI in Decentralized Social Media
(00:05:00) Happy Bitcoin Friday
(00:06:09) Guest Introduction: Alex Gleason
(00:07:02) Truth Social
(00:10:35) Challenges of Centralized vs Decentralized Platforms
(00:14:01) Bridging Platforms
(00:19:13) Limitations and Potential of Mastodon and Bluesky
(00:24:08) The Future of AI and Vibe Coding
(00:31:08) Empowering Developers with AI
(00:38:09) The Impact of AI on Software Development
(00:47:02) Building with Getstacks.dev
(00:53:04) Impact of AI Models
(01:02:01) Monetization and Future of AI Development
(01:14:07) Open Source Development in an AI World
(01:22:17) Data Preservation Using Nostr
Video: https://primal.net/e/nevent1qqs96kxmxc7mufgt6n2rxpphg8ptyx2kl47a7rj389jrwmvjy6rhuhgmfel87
support dispatch: https://citadeldispatch.com/donate
nostr live chat: https://citadeldispatch.com/stream
odell nostr account: https://primal.net/odell
dispatch nostr account: https://primal.net/citadel
youtube: https://www.youtube.com/@CitadelDispatch
podcast: https://serve.podhome.fm/CitadelDispatch
stream sats to the show: https://www.fountain.fm/
rock the badge: https://citadeldispatch.com/shop
join the chat: https://citadeldispatch.com/chat
learn more about me: https://odell.xyz
Hello. Hello. No. Sir, it's part of a long history of people trying to create decentralized social media, including many you've never heard of and some that perhaps you have. Have you heard of activity pub, also called mastodon? Have you heard of at proto, also called blue sky? Or no ster, also called nostr? These are the big three. And each of them have their own unique set of strengths and weaknesses. But let's focus on the bigger picture. Wow. Those are some big numbers for Blue Sky and the Fediverse. But wait, here's the list of top social media websites and Blue Sky is there combined with Activity Pub and Noester.
So after all of these years, why are we still losing? I heard it's because they're funding us too much. Yep. According to egg man ninety eight in the comment section, if they would just cut funding, the sheer desperation would create an innovation like no other. Or perhaps the reason we're losing is because nobody cares. Perhaps nobody cares. That would certainly explain their behavior when they jump ship from centralized platforms over hot button issues only to come crawling back once the dust settles. Another reason people say that we're losing is because of our sign up form because on Nostr, you have to use private keys. But in my opinion, Facebook sign up form has gotta be one of the worst in the world today, and yet they're still the top social media website. I think Noister sign up form is actually its superpower.
Because our biggest competitor is actually not Twitter, it's email. Email, which is supposed to be decentralized in practice is basically just controlled by Google because all of these sites here, most people are just signing in with their Gmail account. But that's okay because Google is good. Right? Now you may be wondering, why not just redecentralize email? Well, the fact is that many have tried, but email is old. Email was created before the World Wide Web, and it was not designed with user freedom in mind. But how will Noester beat email?
With robots. As long as they don't kill us we will harness the power of AI to create things on Noester that people actually find useful. Noester is perfectly suited for it. By the way, I created bridges between No stir and the other decentralized protocols, so I know a thing or two about how they work. And I can tell you that they're not gonna benefit from AI like we will because their integrations are hard and ours are easy. In fact, the robots are already doing it. Here is Chorus, a NoStar community website created with AI in four days. Here is a Scrabble game created in one prompt to AI.
Here is a Notester audio experiment where people can record their voices and be heard all around the world. All of these and more were created with MK Stack, an AI driven template I created for my wife, MK, after I was inspired by her work on Bookster. Here's a site like Bookster created in one prompt to MKStack. The bigger picture is my new project called Stacks. The idea behind Stacks is that developers will create a stack and then community leaders will use the stack to create custom brand new community platforms tailored to their community. All you need is to dream it, believe it, and build it.
AI is the biggest opportunity that we've had since the Internet itself, and it's our biggest chance ever to make people care. All we need is you. Thank you very much.
[00:04:58] ODELL:
Happy Bitcoin fight Friday, freaks. It's your host, Odell, here for another Citadel Dispatch, the interactive live show focused on actionable Bitcoin and Freedom Tech discussion. As always, dispatch is funded purely by our audience with Bitcoin donations. Thank you, Freaks, for supporting the show and keeping it ads and sponsor free. Our largest zap on podcasting two point o apps last week, or I guess earlier this week was from BTC's shelling point ride or die freak. He sent 2,100 sats and said awesome rip. You can join him and other people using the Fountain Podcast wallet, Fountain Podcast app, to zap and comment on every show using Nostr.
Also, of course, you can join us in our interactive live chat that gets streamed into the broadcast and is unmoderated, unedited. You are one of the hosts alongside me. That is also Nostr enabled. You can find all the links at cieldispatch.com. I see we already have doctor Nikamoto in the chat who zapped a thousand sats. Anyway, Freaks, I have a great show lined up for us today. It's actually the man who was in the intro clip, which is rare for me to do, but I just thought his presentation was a fantastic way to get started on today's show. We have good friend, Alex Gleeson here, who has a history building out Truth Social and is now focused on Nostril AI and Bitcoin.
How's it going, Alex?
[00:06:35] Alex Gleason:
Hello. Hello. Happy to be here. Happy Friday, the thirteenth.
[00:06:39] ODELL:
Yes. I thought it was a perfect day to have you on the show.
[00:06:42] Alex Gleason:
Yep. I normally don't go outside. Today, I was planning to, but now I'm a little bit concerned.
[00:06:49] ODELL:
When I when I when I invited Alex on and he realized it was Friday the thirteenth, he sent me a shocked a shocked face emoji, and I knew it was perfect.
[00:07:00] Alex Gleason:
Yep. Good luck, everyone.
[00:07:02] ODELL:
Alex, of all, great presentation in Oslo. I think your presentation style is incredibly unique, and I I, for better or for worse, go to a lot of conferences. So I always enjoy watching one of your talks, and this time I got to enjoy it from afar. Before we jump into Noster or however people wanna pronounce it, I think it's pronounced Noster. I'm a Noster guy myself. But I know you're you know, you can't be right about everything. I wanted to are you comfortable talking about your Truth Social days at all? Absolutely. So Truth Social, what how are you involved in Truth Social?
[00:07:47] Alex Gleason:
Yeah. So I've been building stuff on the fediverse for a long time, many years. And one day, I got a call from a guy saying or or an email saying, hey. Can you help me put Soapbox on top of Mastodon? Soapbox was the software open source software I was developing, for the And so he wouldn't tell me who he was or why, but I had a call with him and he was, like, excited about it. And I was like, okay. I'll take a chance here. And then before I knew it, I was on Donald Trump's social media team building the new website for Donald Trump after he was banned off off of Twitter. And, like, it took me so long to accept that this was actually real, because we were, like, you know, like, hiding around in secret. We weren't allowed to tell people, like, only our closest family members knew, and even that was, like, a gray area. And in the meanwhile, I was, like, flying back and to, to Atlanta and and, like, pushing, like, TVs up the WeWork stairs to do investor presentations and stuff.
And, I have I have fond memories of it. It feels a bit like a dream to me now, because, you know, there were it went through all of the dramas that I think you would expect that an organization like that would go through where, like, at it it was full of people that I loved, and then those people kind of got forced out and replaced with different people who want who had a different vision mainly around, advertising that they wanted to push into truth social. And and so once I realized that, you know, I wasn't gonna be able to do this sort of decentralized thing that that I thought would actually fix that fundamental problem there, then I had to leave and and go on something else. And that's when I jumped ship to Noister.
Because to to me, like like, I wanted to solve the actual issue that arise from Donald Trump being banned off of Twitter. And and I felt like they weren't actually interested in solving it. They were more interested in just spinning up their own, you know, competitor under the same general rules of centralized platforms.
[00:09:56] ODELL:
Yeah. They wanted to be the trusted party and and control the platform. Exactly. Enterprise it. So you were, like, actually one of the leads in the build out. Right?
[00:10:10] Alex Gleason:
Yeah. I was one of the early people there. Because of my experience with Mastodon and the fediverse, that they they had decided to to build off of Mastodon, early on in the process, and then they brought me on. And I was, like, the expert on Mastodon, basically, at that point. So I had a lot of sort of say in the direction of how we would go on the technical side of things, and that led to me becoming the head of engineering at the company. And I was sort of, like, you know, crossing this dream between teams, building the apps on on Android and iOS and on on web. And I was also leading the web interface because the web interface on Truth Social Now is based on Soapbox, the product that that I built, on on Mastodon. So I was, like, sort of helping with with building out new API endpoints and and, like, connecting the teams on it.
[00:11:03] ODELL:
And Soapbox was basically, I mean, how would you describe it? It's like almost like an SDK for building your own Mastodon instances?
[00:11:14] Alex Gleason:
Yeah. It it was like a version of Mastodon that's that's fully customizable. Because one thing I didn't like about Mastodon was they're like, oh, look, you can spin up all of these different communities for your own, like, little thing. You can have one in in, like, for your country. You can have one for your sports team. You can have one for your hobbies and interests or whatever. But but it's all this very strict Mastodon style template where it's this blue, blue gray color scheme and it has the elephant and it always says Mastodon. And you can't change the name Mastodon to say, like, my website or whatever. And I felt like that was a big missed opportunity. So I basically just made, like, a version of Mastodon where you can change the colors, you can change the logo, you can change the name.
It doesn't say Soapbox anywhere in the UI. And so Truth Social saw that as a good starting point for them since they already decided to build off of Mastodon on the back end of, like, us being able to build, you know, the Donald Trump version of Mastodon. Right.
[00:12:16] ODELL:
So, I mean, since then, it's it's become the most valuable Mastodon instance of all time. I think I mean, the stock that owns it, DJT, which is his initials, is worth 5 and a half billion dollars. Do you own any do you own a d j t stock?
[00:12:34] Alex Gleason:
I don't, unfortunately. I don't think that anyone really does that that was promised that they would that they would. They kind of they got cut out of the deal. You're fired. So it's about what you would expect from a Donald Trump organization, you know.
[00:12:50] ODELL:
Yeah. Anyway, okay. Well, thanks for giving us the history lesson there. I think it's fascinating. I, before we continue, huge shout out to Mav 21, rider die freaks, and 10,000 sats. Pau Pauley sent 2,100 sats. Thank you guys for supporting the show. I think it's, one thing he does deserve a bit of credit for is that he is pretty militant about only posting to truth social.
[00:13:22] Alex Gleason:
That's true.
[00:13:23] ODELL:
I thought I mean, I think it it was a fascinating, it was a fascinating case study on the the current status of social in society, the Elon versus Trump war, where Elon was tweeting on x and Trump was responding on truth social and this is a cross platform like, a cross platform war.
[00:13:52] Alex Gleason:
Damn. Isn't it a shame that they can't be linked together by some sort of protocol? That sucks.
[00:13:59] ODELL:
Well, that's what we'll be talking about today. So, I mean, I think an interesting part here is so let's dive in real quick on what you because you're it's it's a weird dynamic where it's almost been harder for me to convince someone who has already switched to Mastodon to try out Nasr than someone who has never never used Mastodon. Like, they're they're just most they they have never used a federated protocol. They've only used a completely centralized platform. It's a good point. So what what what made you move? Like, what are the specifics about the differences between like, what were the failings of Mastodon that you think Nasr, fixes or mitigates Yeah. So
[00:14:47] Alex Gleason:
So it it's futile, basically, on Mastodon. You have, like, lords landlords, basically. And so you have to kind of buy into this system of being a renter of, like, a apartment complex as a user. And and, like, that that's not really freedom to me. In in many cases, that's actually worse because at least with big centralized, like, big tech evil platforms, you kinda know what you're getting yourself into. And probably the the people in charge of that aren't gonna be coming after you on a personal level. But on Mastodon, like, the way it works is you have to you have to choose a server. Right? And this this is already, like, a major issue because it's like it's like the psych psychology thing of there being too many items on the menu and that making people not be able to choose.
And and so you you choose. And then once you choose, it's permanent. You're stuck there. And somebody else owns that too. You don't own that. So you're just basically attaching yourself to somebody else's, like, property. And then you your account is just at the whims of that person forever. And you there are, like, there are ways to move your account, but they're really, like, an afterthought and not deeply integrated into their protocol, and they don't work very well.
[00:16:09] ODELL:
So And the key is you you lose your social graph in that situation.
[00:16:14] Alex Gleason:
Exactly. You lose everything. And and and, like and you you you need permission to move too. Yeah. So so so, like, if you if you're in an adversarial relationship with your landlord, which is very likely in the Mastodon, situation because, they're not sending their best. Like, the the people who are running the Mastodon servers are just like, you would hope that they're, like, principled, like, freedom advocates, but they're not. They're, like, people who are are emotional. They're people who are, like, power tripping. And so you get in these situations where it's, like, it's damaging to to this decentralized mission and future because people just get such a bad taste in their mouth. They're like, oh, I never wanna do this again. Like, honestly, centralized is better.
So it sucks. But for the people who don't have that experience so I will say if you're running your own Mastodon server, then you're not gonna have this issue because you are fully in control. It's just, like, most people aren't gonna do that. That's not gonna the mainstream is never gonna be able to do that. So so those people are are happy. And then for the for the people who are just fine sort of being beta orbiters of these people who are running the Mastodon servers, then those people are also fine too because they're just like, yes, daddy. Like, I want you to be my admin and tell me what I'm allowed and not allowed to do on social media. And and that's fine for those people too.
But, you know, with Nostr, it or sorry. With Noister, it really gives people full control over their their identity online. And and that is so important be because, like, we've seen these issues on Mastodon where where, like, admins will just get get fed up and just delete their whole server and then, like, thousands of people are just, like, gone. And that sucks when you invest in something. You you've, like, built up your follower base, you know. You've you've, like, posted all of your content to it, and then it's just, like, someone else can just press a button to delete you off of the Internet. So, I it's not a winning strategy in the long term, and so we need something like Noister.
And and I think Blue Sky has, you know is the same issues as Mastodon too and it's just that they're a little bit of better stewards about it. So people feel comfortable being in this sort of middle ground where they're on a thing like Mastodon, but the the admins aren't just, like, the worst people ever. So, yeah. But but but I think that but now they're facing kind of bigger issues with, like, governments and stuff, pressuring them to do things just like Twitter was. So so, like, I think Nostra is the only actual solution to this problem that we have right now, and we just need to get people to understand that or, you know, just wait and then something will happen that will cause them to understand it.
[00:19:14] ODELL:
I mean, and the other piece too is is with with Mastodon is you're you're not only, it's it's not you're not only beholden to malicious instance operators, but also, like, something could happen to like, it's it's it's very burdensome to run instance. Oh, yeah. And I think there was a story where there was, like, a
[00:19:41] Alex Gleason:
a gay focused instance and, like, their domain got cut and then just everyone got absolutely screwed. It it masked it on it it was, like oh, I don't I don't know if this was the same one, but but there was a .af. It was, like, masked on that a f or something that was actually, like, the acronym, like, as fuck. Right? But but it's actually a f is owned by Afghanistan. And so, like, the the Taliban run government in Afghanistan was able to take down that server with, like, thousands of users on it because they they chose the wrong top level domain provider.
[00:20:16] ODELL:
Yeah. That's what it was. It was it was gay.af,
[00:20:19] Alex Gleason:
I think. Yeah. Yeah. It was something like that.
[00:20:23] ODELL:
Yeah. So, I mean, that's not a scalable situation to be in. That's just, it just doesn't make sense. It's fundamentally flawed. And then so let's just really quickly I mean, you've been one of the things you started to do was, build bridges. Right? Because this is something that we hear a lot on people who are joining Nasr for the time. Let me just see, you know, the most interesting Twitter accounts or let me like, I mean, one of the most useful accounts that I find on Nasr is, the mirror that you have of Donald Trump's Truth Social. And so you started to try and build bridges. And, obviously, on closed platforms, they can stop you from building bridges, which just inherently proves that they're flawed.
Yep. I mean, they're not actually decentralized or distributed or whatever you wanna call it. But so you've been trying to build bridges to Blue Sky. I know you got, like, some of my posts on Blue Sky for a bit. I think it's since broken. So you're kind of familiar with how Blue Sky works in their ad proto. What what are the issues with that? Like, as someone who's, like, not really familiar with how Blue Sky works except that it just seems like it's Yeah. So from the beginning.
[00:21:34] Alex Gleason:
Yeah. So so I I will say with Mastodon and BlueSky, they're at least incrementally better than the centralized platforms because you can at least get stuff out of it. On the centralized platform, it's like like so you need to register for a developer account, and then we might give you an API key, and then we might charge you money to use that API key. And and aside from that path, you're completely out of luck. And and also one of the main things that you need to create a bridge is you need the ability to generate infinite accounts on the other side of the bridge. Because, like like, we have Noister users and we want to make them appear on the other side. Right? So so, like, on Mastodon, you can you got you actually can do that. You can have a server, most of that pub. Right? That can just have a dynamic,
[00:22:20] ODELL:
route for user Yeah. I saw the Graphine account on Mastodon responded to me. I didn't know they could see my responses.
[00:22:27] Alex Gleason:
Yeah. They can. They can. Yeah. And That's pretty cool. And and on on Blue Sky, it's a little bit more challenging because they have this system where they have this centralized server that in theory could be replaced. But in practice is what everything is running through. I I think that their philosophy is kind of like, the communist like, Marxist Leninist idea of the withering away of the state where, like, like, the theory is supposed to be that the communist party is going to take control of the state and then they're going to become the new state. But then they're gonna do everything the right way and then over time, the state is not gonna be needed anymore. Unfortunately, we've never seen that actually play out in practice.
And and I it seems to be that that Blue Sky is taking the same approach here where they're they're just like, okay. We're just gonna take power and take control and then make everything right, and then the world isn't gonna need us anymore. So, so yeah. So we have we have bridges a a bridge to Blue Sky that is kind of functional. We can get data out. Getting data in is a lot harder because they expose all of their posts publicly. We can we can get stuff out of it. But, in terms of why it's broken right now, it's just because of the cost of hardware, and the amount of time that I would take that that it would take in order to fix it.
And and so now now I have some, you know, some resources to be able to fix it, and I just need to put some time into it, because it's really complicated the way that they've built things. And and on Noister, it's it's a lot easier. So I I don't know
[00:24:13] ODELL:
You could do all of that and then they could just block, like, 99% of users from seeing any posts. Right? Exactly. So so I I was gonna say it's also, like, a matter of
[00:24:23] Alex Gleason:
how much time is it worth spending versus the the time I would spend growing Nostril. And and I'm kind of a little bit more bullish on the idea of growing Nostril than fighting with trying to make Blue Sky work and, like, begging for their permission at various steps of the way.
[00:24:40] ODELL:
Okay. Yeah. That makes sense to me. Fuck blue sky. Anyway, okay. So we did that. Now I I think it would be actually, pretty helpful to people just as a thought exercise. If we want let's just say we wanted to mirror Elon Musk's x posts to Noster. Why is that a bad idea? Why like, how how difficult would that be?
[00:25:07] Alex Gleason:
I think that it's possible to mirror one, sort of, like, how I'm mirroring one Truth Social. You know, Truth Social, even though it's based on Mastodon, they've removed all of the sort of federated features of it. So it's also basically like a centralized platform at this point. It's also basically the same issue as, as x. Maybe a little bit less bad. But but, like, I for the same reason, I'm not gonna mirror all of truth social. I I think that you can't really do more than, you know, a couple of x accounts because you you're gonna start hitting rate limits. You're gonna start, like, they're they're gonna start detecting your activity and blocking you. And then potentially, you're also running into legal issues for, like, violating the terms of service. And then, like, god forbid, the worst case scenario is you somehow get charged under under, like, the computer whatever whatever law, the one that Aaron Schwartz got hit with. And then and they're, like, accuse you of being a computer hacker accessing unauthorized systems or something.
And it's just like, god, like, you know, actually, maybe it's just regular people making the choice every day to continue using this that's the problem. Like, maybe y'all should just stop stop giving these platforms power and and then, like, and then these the then the issue would be solved, you know. Like, why is it why do we have to take all these big risks, when it should just be an an act of realizing that that it's wrong and then convincing everyone around you to just move on to the thing that's right. You know? Yep.
[00:26:45] ODELL:
Yeah. So, I mean, by the way, do you, is your volume really high on your headphones? I hear a little bit of I hear myself I hear myself every once in a while.
[00:26:58] Alex Gleason:
Yeah. I'll turn it down a little bit. Okay.
[00:27:01] ODELL:
Perfect. I think that's better now. Cool. But I presume you can still hear me. I can hear you. Yeah. You're good. I, so, I mean, in practice, it would be like we'd have to run a bot that's, like, manually scraping from the website rather than pulling from an API.
[00:27:21] Alex Gleason:
And potentially from multiple IPs too.
[00:27:24] ODELL:
Right. Because they would see the activity and then there it'd be whack a mole, it'd be unreliable, and probably just more trouble than it's worth. I mean, you can also just always do the manual practice too where it's just like some dude is sitting there just, like, reading the web page, typing out the typing up. But it's ridiculous. I mean, that's that's this is the whole reason Nostra exist Is that is that the whole the the majority of the Internet is going behind walled gardens. The thing Elon did when he bought Twitter was put up a wall and make the wall higher and higher, require logins to to view content, lock down the API.
And one of the main reasons for this is because they see the data as their product. They see, user posts as their product, and particularly in AI world, it's it's data that they can train and use in real time. One of the reasons XAI or Grok, is is so useful to people is because it has basically, of a fleet of human robots that are just feeding information into the sys into their proprietary walled garden that only they can easily access, that the other AIs aren't able to easily access and and leverage off of. And Nastr will hopefully fix this if we can scale it up and get more people involved. So, anyway, with all that said, I thought your talk was fascinating. You've recently been more focused on this AI intersection.
Why and, why should like, why why do you care? Why, you know, why should people care about this?
[00:29:10] Alex Gleason:
Yeah. So so the story is, you know, for the past, like, three years, I've been trying really hard to make no strings succeed, and I've been trying to solve really hard problems. Like, every day I'm banging my head against the wall, basically. You know, just trying to create, like, the the ideal technical solutions to these things and trying to, like, plan these, like, long term of, like, this has this can't just work for today. This has to work ten years from now. And and and then, like, you know, three months ago, my wife started playing around with with AI, and she built a Nostra site in, like, four days called Bookster.
And and that just blew my fucking minds. I'm like, and you didn't even need me to build that? You just built that, like, completely on your own? And and I'm I'm, like, shattered, you know. I'm, like, having this major existential crisis. It's like, oh my god. The world doesn't need Alex Gleeson anymore. What am I gonna do with my life? And and so I'm, like, I have to confront this because, like, I wanna be useful. And it's it's it's honestly scary how fast things have moved in such a short period of time. And so I'm like, okay. AI or die. Like, that's it. Like, you know, I feel like we're in this this this situation now, this kind of, like, and and I'm I'm a little bit less concerned about it now that I've come back to reality. I'm like, okay. Everything is fine. Everything's gonna be okay. But, like, I think it it's it's kind of like the Internet was, back in the day where, like, you have to adopt this thing or, like, you're out because you're just not gonna be able to move as fast as the people who are because it's just such a world changing, game changing difference.
And and I I think that we can do things that are extremely powerful with it to make no stir win. And so and and it doesn't have to be so hard anymore. You know, I I built one more hard thing. I built stacks and I created MK Stack. And a lot of people have been using it now. And it's been a lot of fun. I've been having more fun lately than than I've had in years. And and there's just there's so much we can do with it now that we couldn't do before. Like, we had this big dream of Nostra years ago that we're gonna integrate Nostra into everything. Like, Nostra is gonna do the other stuff. Nostra is gonna is, and it's all gonna be linked together and but but it was just hard.
And now with AI, we can actually build all that stuff. Like, you know, I I remember years ago, Jack had had mentioned he wants a GitHub on Noister. And people worked really hard to get out a spec for that, and people worked hard to create these sort of, you know, Gnostr Git platforms. And and and so using that nip 34 spec, I I just was like, hey. Build me a nip 34 client to my MK stack template that I built. And now we have nostrhub.il. And that's a site that I spent about four days working on where you can on that site, you can create custom NIPS. You can submit your nostril apps. You can submit repositories. There's a DVM marketplace. There's a NIP 72 community built into it. All this shit is, like, integrated and deeply linked together, and this is all just a vibe coded app that is fully functional.
And and, like like like, every passing thought or idea that I have at this point that like, years ago, I'm like, oh, no. That's too hard. That sucks. I can't do that. I'm just like, hey. Build this thing. And it just builds it and it works. And and and I think that, like I said in my speech, NoSir is better suited for this than anything else because, like like, other people have these tools that they're using similar to this, but they're struggling to get a login button. They they really are. They're they're like they're like, you know, build me this website or whatever and then it's full of mock data. And it looks like it should work, but it doesn't work. And that's because the thing that they're missing is Noister, because plugging it into, like, the these legacy web two systems, requires a back end and it requires all this extra work. But but Noester's design is this, like, weird alien design that sort of switches the roles of front end and back end where it, like, wants the front end it wants smart clients and dumb relays, and the relays are permissionless and publicly accessible without authentication. And you can just generate IDs on the client side, user IDs on the client side with keys.
So, like, that's what MK stack is, is I've basically taken the lovable stack and added Noister to it and then put a context file in there. And now we have this machine that can just generate and pump out Noister apps. And this past month, I've been going in there and just refining the context file, adding little snippets of code in there to do things that we wanna do such as upload files to Blossom. So now there's just like a use Blossom hook in there that that can be called by the AI, and and there's just a little bit of context in the context file telling it how to do it. And so now you can be like, and let me do uploads, and it just works on the shot. And so, the way I've been iterating on this thing on MKStack is is this template that's a React, Shad c n, v t, Tailwinds template.
And I just be like, build me a site like Twitter on Noister, and then it does it. And I see what it did wrong. And then whatever it did that I didn't like, I go back into the context file and I say, like, uploads, for example, uploading a file. To upload files, use this snippet of code right here, like and import this module. And so then I run build me a site like Twitter and make it so people can upload files. And then the time, it does it. And so what I'm like, my goal here is that you can write a single prompt, and it will build you what you want in the shot without needing to do any additional, like, request for changes.
And and, and so we started off just doing, like, existing NIPS. Right? Like, build a site like Twitter or NIP 28. Build me a NIP 28 chat applications. These things all work. They work on the try. And now we're starting to move into the territory of, build me something that doesn't yet exist. Build me something that there's not a nip for. And so, when it has to do that, it it does a lot of things wrong. It it tries to use tags wrong. It queries things wrong. And so I've started to go back in and be like, okay. Build me a Noister platform for the farming community. And so it what it wants to do when I say that is it wants there to be a marketplace where people can buy and sell tractors.
It wants there to be, like, it wants, there to be a feed that shows people weather conditions that are changing over time and for there to be, like, a farming, community in there. And so, I'm kinda just go along with what it wants to do, and I and I'm just go back into the context file, and I'm teaching it. Here's how to write Noester events the the correct way so that you can actually create a NIP for this. And now now when you say build me a Noester farming community site, it'll go and it'll it'll look at the nips and it'll say, okay. I see a NIP 15 marketplace and I see a NIP 99 classified. NIP 99 classified looks closer to what I want, but I need to make sure that I t tag it with farming so that we're only getting farming classifieds. And it just does all of that automatically now, and then it creates a nip.md document explaining that it's using nip 99 with with t tags that say farming.
And and it it also has the ability to generate new event kinds and then document those event kinds. And then my goal is that there will be a simple command to also publish your NIP MD to Nostra hub so that other people who want to implement that same NIP can just refer to it by its Str address and say, you know, implement a No Str farming website using this NIP that this other person made on their Vibe Coded farming website.
[00:37:30] ODELL:
So there's not a NIP. It'll create a NIP. So then it'll be easier for other people. I mean, the key here the key here is when you do this stuff on open protocols, you just it's it's the lack of permission. There's you don't need to ask any permission of anybody. And you have Nasr handle the identity, the comms, and the the data. And then you have Bitcoin handle the payments, and then boom, you're off to the races without asking permission. Yep. So so let's just I guess, let's go high level and then we'll go deeper.
High level, you know, there's some people have been more provocative than others in terms of what what this new era of of quote, unquote vibe coding means. Victor, who, has manually coded Amethyst, one of the largest, most popular Android, Nostra apps, has gotten very excited about this even though he doesn't really seem to be practicing what he preaches yet. But he thinks he's about to be obsoleted by Vibe coded apps. Obvious like, the three largest I think the maybe the four largest nostra apps, if you do Domus Primal, Amethyst, Jacajohn, maybe those are the four largest. Maybe they're not. But, anyway, none of them are vibe coded.
When I look at it from just, like, a high level, do you do you think, like is there I could see how interesting tools could be built very quickly and iterated on very quickly with this type of, technology, these this type of tool set. I mean, Nostra Hub, like, looks like a fucking legit legit website. Like, it put very performant. It it works exactly like you would think it would work. But is there a concern there in terms of, like, long term maintenance, like, actually long term stable projects versus, like, kind of just, like, not I'm not trying to be denigrating here, but, like, hobbyist tools, you know, like, just,
[00:40:01] Alex Gleason:
Yeah. Absolutely. I I mean, like, as the project grows, it's for sure gonna just get crushed under its own weight, but that's also kinda the same issue with, like, the, the manual code. I call it Discord coding, by the way, because Discord is the opposite of Vibe. So, you know, the discord coded apps also have many of the same issues where, like, as it grows, it starts to get crushed under its own weight. And one of the beautiful things about Noister is that it really like, due to the way identity works, it allows you to build many small, like, micro apps. And and this was something that people were so excited about in the earlier days of Noester, and I I just feel like it didn't really play out that well because it was so hard to do it. It's like, oh, you have these random ideas of all these various different things you wanna do, but is it really worth sitting there and taking the time to, like, code all this stuff out? And most likely, it's not gonna work very well. But now that we have the ability to buy code, we can just do all of that. And we can build these, like, thousands of micro apps that are interconnected, and they just work. Well, let me put it a different way. Yeah, that seems
[00:41:07] ODELL:
incredibly empowering and powerful, and I'm very excited about it. Let me put it a different way. Do you think like, there's one thing I see a random m bump in the comments. Damn. Not used to hearing Odell speaking at one x. Yeah. I've heard that in person a lot. I, feel free to come back and listen at two x later. I, it's one thing if I'm building, like, a chess app. I one shot a chess app. You play with people in chess. People don't really have to there's not, like, significant technical burden there. There's not significant user expectations. But do you think in five years or six years that the main Twitter like app on Nasr, Right? Or, like, there's, like, the almost, you know, like, the the one that is being used by the most people. Is that a one shotted is that a one shotted vibe code app, or is that a, you know, like, a heavily engineered maintained with So You know, still using LLM tools. Like, you're still using a AI tools in the actual development process and the maintenance process, but it it involves a team with humans, like, actually interacting with it on a constant basis and iterating and creating a stable, improving experience.
[00:42:32] Alex Gleason:
So I won't try to predict five years. Things are changing so much, like, every week. But I today, I think that we could build a Twitter app like that. Would it be one shot? No. You one shot would get you something that looks really good and gives you a feed and has the right buttons on it, but you're gonna have to tell it, make the like button work, make replying work, fix the mobile padding, you know, add this feature, fix that, like, add bookmarks, add all these different things. And then it will eventually get to a point where it starts to get kind of weighty, and it you might start to see the limitations. But but I think that, like, the way that the new way I'm approaching software development at this point is not to try to go in and discord code these features into these, like, five coded apps. It's just see what are ways that I can improve my stack so that when I rerun these prompts, it'll produce a better result the time and it'll do it faster. And some of that does involve going in and manually, like, taking pieces of code and and moving that code into the stack template itself so that it just already has that code in there.
And through that sort of software development flow, which which to me, this is like just a it's a completely new software development methodology. Like, when things get stale, you may consider just deleting and reviving because, like like, I I I kind of view it as, like, every six months, you just revibe the app. And, like, you you're improving your stack the whole time and upgrading all of your shit and improving the components in there.
[00:44:16] ODELL:
I could see that's how that that's power like, incredibly powerful. Right? Because you don't lose anything in terms of, like, the actual database that's being used, the social connections, the identities. Like, that's all separate and permissionless, obviously. But do you think users would be cool with that? Like, they like, I'm just like, I'm not talking about, like, technical users, which is basically all we have on Nasr right now. Users. I'm talking about, like, average Joe Boomer or average Joe Zoomer, like, signs into, you know, opens Amethyst on their phone, and it's just a completely revision stack or whatever. Is is that, like, something that you you think users get around? Is that is that a I don't know. It's a weird thing to grab to grapple with in your head.
[00:45:08] Alex Gleason:
I'm focusing on the web, and a beautiful thing about the web is that, like and and these are static files too. So one thing that we have been able to do recently is make it so you can vibe code your app your app with MKStack and then publish it to Blossom. And and then people can just go onto a Blossom URL and just use the app because it's all static files. Right. And and so, like, we can publish multiple different versions. We can have the like, the version from six months ago can stay there essentially forever, And you can just download those files and you can run them for as long as you want forever. So if you really are To be clear
[00:45:44] ODELL:
to users, a Blossom is a nostr spec for data files, whether that's media or other types of data that is hashed and signed, so that you can make sure that it has integrity, that it hasn't been altered or changed, and you can mirror it between servers. So in that situation, you could actually host that stuff on Blossom and then not even have to rely on
[00:46:08] Alex Gleason:
domains and DNS and all that stuff. Right? Exactly. The whole the whole app is on Blossom. It's just index HTML file with CSS and and JavaScript, basically. That's fucking awesome. Yeah. So so people can you know, we'll we'll have, like, way back machines of of apps where we'll be able to see, like, you know you know, Nostra Hub in 2025, Nostra Hub in 2026, Nostra Hub in 2027, and and, like, people can still use the old versions if they want.
[00:46:36] ODELL:
I could use, like, v 23 if I'm attached to it. Exactly.
[00:46:40] Alex Gleason:
So users are in control here.
[00:46:44] ODELL:
Interesting. Okay. I don't know where I wanna go next. Okay. So get stacks.dev is the project, website. Right?
[00:46:53] Alex Gleason:
Yeah. People have been very confused about stacks versus MKStack, by the way. I know. It's a hard as a Big Corner,
[00:47:02] ODELL:
horrible name. The shit corners ruined, the low
[00:47:06] Alex Gleason:
stacks. I noticed I've been fighting with with some shit corners on, like, domains and stuff.
[00:47:11] ODELL:
Maybe we could maybe we could take it back. I mean, I will say when I made Sats the standard, we couldn't get any good domains
[00:47:18] Alex Gleason:
because the SATs. So, you know, it
[00:47:21] ODELL:
it it is what it is. Maybe we could take it back. I think we could probably just take it back. It's an irrelevant shit coin, and most people don't know what it is. At least you didn't call it That's my feeling. At least you didn't call it Ripple. So get stacks.dev, I just signed in, with my nostril extension. So, like, what is the process for someone to actually
[00:47:47] Alex Gleason:
can we go through what the process is for someone to actually build something using this? So so right now, you have to do it in a terminal. And that's something that we're working on changing right now. But the process is basically there's an install script that you run. You have to have Node installed already on your computer, and then you use the NPM install to get the stacks commands. And then you'll just have the stacks commands. And then you have to choose an AI provider. So some options include open router, paper queue, and router. All of these work. And once you've decided on that, then you you can clone the MKStack template, and then you can just and MKStack already comes pre configured out of the box with support for for Noister.
So you can just say build me a site like Twitter, and it'll automatically build that on Noester without you even having to say the word Noester. And and so the the idea is And so what does and then what does it spit out? And it spits out yeah. So so then so it creates a new folder on your computer. And and that folder contains the template files of MKStack, which already has a lot of things set up, but it's just a blank website. And then it goes and it edits them and it uses components that are there to sort of assemble them on the page, including, like, the login button. And and then it'll go and it'll style it. It'll install custom fonts on there depending on like, if you're farmster, then it's gonna give you, like, a a something that looks like, you know, eco green farm type of vibe.
And it's gonna it's gonna style that CSS for you automatically, and it's gonna put, you know, the components in there to make something that's functional. And then it'll be like, alright. I'm done. I finished it. And then you you open a new tab and you type npm run dev, and it'll give you, like, a local host URL you can click on and then you can see Farmster. You can start clicking around and log in. You can you can do stuff. It's real. And then if you wanna you wanna share it with people, then you can run npm run deploy, and that will upload the built files to Blossom. And that'll give you a public URL that you can share with people and say, here's, hey, guys. I just buy a code at Farmster. Check it out.
[00:49:58] ODELL:
What is that public URL? Is that you just are hosting that?
[00:50:03] Alex Gleason:
So we're hosting no nosternosterdeploy.com. Okay. And that and and so it's it has a sub it's using the Noister static, or the the Blossom static website specification where there's a n pub in the subdomain. And and so it generates an n pub per project and then uploads to that to that end pub. My that's not my end pub. That's No. It's a project generated end pub. Yeah. And then and then other people can access that through the bulk of the actually like it, then I could go through the manual process of registering a domain and hosting it myself, blah blah blah. Exactly. And we we have, like, scripts in there to make it easy to use GitHub pages or GitLab pages as well. So if you just if you push the code to GitHub, then then you just have to go through, like, a simple process to get it on on git GitHub pages. And then it's free. It's free either way, basically. The only expense is through is through the AI provider.
[00:50:59] ODELL:
So does it matter which AI model you use?
[00:51:03] Alex Gleason:
Yes. I mean, it must. Claude Claude is the best one in my opinion. People will fight me on that. They'll say Gemini or some people are, like, you know, big fans of GPT four. But but Claude is the one that's the best at tool use. And, you know, we we have options. But honestly, the thing that sucks about it most for me is that, like, these AI companies are the new big tech. Like, I I kinda think Google and Facebook are out the window right now, except Google is staying relevant by by having Gemini. And that's the you know, we're gonna be able to just vibe code the next Google or or Facebook. But but we what we what normal people can't do is is train and host AI models.
So that's, like, our biggest
[00:51:51] ODELL:
threat to freedom in the future. I And so what what's the best open source model right now? Is it LAMA or is it or is it DeepSeek? Or is it something else that I'm like Neither
[00:52:02] Alex Gleason:
I've tried both of them and neither of them worked very well for me on these projects because they're, like so Claude is made by Anthropic and and Anthropic invented the the concept of tool use. I'm I'm sure they're not the to try to do it, but they're they're the ones who made it actually good. And so tool use is what allows the AI to do things like read and write files on your computer, and to, like, run commands on your computer. MKStack needs in order to be able to, like, actually edit a code base and make it functional. Because because it's not just, like, editing files. It's actually running commands to it to get feedback. So it'll it'll start to, like, add and edit files, and then it'll run tests automatically. And then the test will give it feedback and tell it you've got type errors on these lines. You need to go fix those. And so, that's how you get a working app on the try is is that it does these these things, and it has this feedback loop. So none of the other models have really caught up to Claude in terms of tool use yet.
[00:53:05] ODELL:
And you can use Claude through something like paper queue and then you'd not you don't have to KYC or anything. You're just Exactly. You pay you pay with Bitcoin. Right?
[00:53:15] Alex Gleason:
Yeah. Or or with Routster. And and those are both exciting options to me because because they they allow people to do this anonymously and to pay with Bitcoin, but it's kind of a band aid over the problem rather than solving the underlying problem. And so I think that's gonna be the next sort of bastion of freedom online is is the front lines, I should say, is trying to figure out how to combat these big AI companies. Because they're basically just like a layer on top of, like, the foundational model. Like, the actual model is still closed and prepared. They're kind of like, you'll you'll interface with us so that your experience will be less bad, but we still have to interface with them.
[00:54:01] ODELL:
Got it. Fascinating. I mean, so where do you see this all going? Are you is the goal to make it even easier than that?
[00:54:14] Alex Gleason:
Yeah. I mean, I think that that what it means to be a developer is gonna fundamentally change. And I I I I you know, people will fight against this, but but, like, what it means to be an artist has already changed. And and people people are already saying if you use Photoshop, then you're not a real artist for years now. But Photoshop is just a tool. And, like, you know, is it the actual skill of the hands or is it, like, the ideas and the thoughts that make you an artist? And and so it's sort of the same question now with code is, like, you know, is it is it your actual skills at being able to do the, you know, functional thing or is it sort of the ideas that you have?
And I I went to, in in Oslo, I went to the to the Munch Museum. Is that how you say it? Munch? Munch? Oh, no. Yeah. The Scream guy. He has the, and we saw The Scream and stuff. And and I was just thinking, you know, like, these guys in 2025 would probably be VibeCoders. These, like like, Van Gogh in 2025 probably would be a VibeCoder. And and probably would be making electronic music too. But, you know, nevertheless, I I I I think that there is an opportunity here to to change what it means to be a developer. And I'm really interested in that because I think a lot of people have great ideas that are and they're just held back by the fact that they don't know how to do it.
And with Noister, we're gonna offer them the easiest path to do it that no one else can offer. So I'm excited to see, you know, things like what Derek Ross, has built are really exciting. And and I think that there there's opportunity for for people who are just like You just name dropping him because he's in the chat? Yeah. But but but also, like, you know, I'm just excited by the stuff that he's building and and and he even even Jack has started building stuff with AI now. And I think he hasn't really coded stuff in many years. But he his his head is full of ideas.
And Right. And and so, you know, it's it's just it's I feel like we've we've been getting kind of stale with it just being these traditional developers building stuff. Right. And and and in many cases, traditional developers are stuck in the mindset that everything is hard and everything has to be hard. Whereas, like, normies are more wanting things to be easy, and so they're they're not telling the AI make it with this or that framework. They don't necessarily care. They just want a particular user experience, and so they're describing the user experience, and they're letting the AI take control, and they're just riding the wave. And and that seems to actually produce better results.
[00:57:10] ODELL:
Yeah. Fascinating. I had a question. Oh, so you guys have you you mentioned Jack. So, like, you, Jack, Raval, Cali, and Jeff announced a new project and other stuff Yep. That is focused on this, that is focused on, you know, vibe coding a better world. How would you, like, how would you explain that project? Like, what's the goal there?
[00:57:44] Alex Gleason:
We're still kind of figuring it out. But but for me, my perception of it is exactly what we've talked about today. And it you know, this idea of other stuff that we've had for so long and no stranger never been able to fully realize, I think with the power of AI, today it can be realized. And I think that, you know, it it's almost like we've we've given up on this idea because it's been so hard for so long that it's like, oh, I don't know. Like, we've tried everything and nothing is is working. But but now, like, now the game has changed.
So other stuff, like, you know, can be viable. I think it's it's it's even a winning strategy.
[00:58:28] ODELL:
Do you think it'd be is it fair to describe it as, like, a skunk works five coding lab, basically?
[00:58:38] Alex Gleason:
So that's definitely how it was when we were building Chorus together, in the in the Alps of Switzerland. And and in in terms of the long term, it's it's kind of like we wanna build sort of, you know, easy onboarding stuff for for people to be able to get into Noister, but I think that we're gonna do that through VibeCoding. And and there's also some big picture problems we're trying to solve too. Like, Jeff is working on, getting MLS stuff. We're working on the GhostR.
[00:59:13] ODELL:
Okay. So I'm gonna ask you a stupid question. Like, why hasn't he one shot at good DMs for us yet?
[00:59:18] Alex Gleason:
Yeah. So so there like, there's definitely still boundaries. Right? And and so there's things that require humans to push boundaries still. And and that's one of them. Because, like like, we're not gonna vibe code MLS. Like, Jeff and his team are gonna discord code MLS and then make a library in TypeScript, I think. The the then we just have a context file in the AI that says, here's how you use the MLS library, and we try to simplify that and make it so that it only has to do the most simple things. Ideally, the I the AI is just a designer that places components on the page, and those components just work already because humans have made them work. And and it's a similar story with Cali and Cashew.
As we were building Chorus, it required a lot of manual Discord coding on his part to get the Cashew wallet to work because the AI just doesn't know how to do these hard things. And that's why, like, you know, my contribution to that project was bringing that template, the MKStack template along. And so that's why the template is so valuable, I think, is because you can collect these sort of discord coded pieces of functionality that the AI is not good at and put them into a template and then tell the robot, you know, you can just include this one, file and then put this one piece of code somewhere, and then it's just gonna work out of the box. And don't worry about trying to implement a Cashew wallet from scratch. Just worry about where you're gonna put it in the UI.
[01:00:48] ODELL:
If that makes sense. So, like, you need humans to kinda create the the fundamental building blocks at
[01:00:56] Alex Gleason:
Yes.
[01:00:58] ODELL:
Which is, like, the same case with Nasr and Bitcoin that is getting then put in to the the toolset.
[01:01:05] Alex Gleason:
Exactly.
[01:01:07] ODELL:
So, I mean, that kinda answers my next question, but I'm gonna ask it anyway, because I'm kinda curious for my own projects. Jack has also, contributed $10,000,000 into this project. Like, what do you guys need the money for if it's so, like, what is the money gonna be used for if you're just one shotting things left and right?
[01:01:31] Alex Gleason:
So yeah. So we're hiring people. The goal is to kind of bring on people onto these pillars that involve sort of fundamental things about, you know, building Nasr clients. And, yeah, like, humans still for sure are are gonna need to be involved. Right. It's and but, like, for me, my biggest mission is sort of bridging the gap between AI and, like, the Nostra community. I want, like, all of the people who have ideas about stuff to be able to just do things on their own without feeling like they can't. I, like, I want people to be empowered. And so I see our role as being able to bridge that gap.
And and some of it does involve doing these these hard things that still need to be done such as, you know, building out MLS and and Cashew.
[01:02:27] ODELL:
Yeah. I mean, so, like, Ravel, another, you know, legendary contributor to Noster. It is amazing how few users we have, but our contributors are so strong as hell. He was on he did an interview with Nifty from Oslo, and he said, and it's not a direct quote. He said something like, you know, 10,000,000 doesn't go as far as you think it does. Am I not wrong to think, like, it goes way farther now that because of this because of this tooling?
[01:03:00] Alex Gleason:
I mean, so much has changed in the past three months, that I think that it's definitely changing how we're thinking about approaching the situation. Because definitely, like, when we started talking about the situation, I didn't even know as as much as I know about AI now. And so Right. Like, I I feel like that's a problem with everybody right now. Like like, I have people on my team who are, like, junior developers, and we're trying to figure out what's your career path now. Right?
[01:03:31] ODELL:
Because, like What is it? Like, was that a har was that a horrible advice when it went viral, like, five years ago for people to learn to code? Or
[01:03:38] Alex Gleason:
I don't think so. But but but I think that right now, things are changing to where, like, like like so I have this mentorship team. Right? It's like young developers, who I'm sort of training to, to do code. And my this started, like, a a year or two ago. And and my approach to this is, like, okay. I'm mister Miyagi. Right? Like like, wipe on, wipe wipe off or whatever. And and that's what I'm doing the whole time. And I'm trying to, like, I'm trying to take them on the graph from, like, senior or, like, junior developer to, like, mid develop middle developer to to to senior developer.
And sometimes I'm drawing the line to middle, and sometimes I'm trying to draw the line straight to senior and just, you know, push them hard to see if they can go. But but now, I'm not really trying to draw the line anymore. Like, I I feel like it's kind of pointless to go from junior to middle or, like, to senior be because, like, if anything, going from junior to middle seems to corrupt the minds to where it makes people when they try to prompt the AI, they start to have opinions about what they want the AI to do in this or that framework or technology or language. There was no too much. Yeah. Exactly. And and so that that actually seems to like it's like the IQ bell curve meme of, like, the the smart and the dumb are are, like, aligned, basically. And once you go into the middle, then you're, like, you're doomed because, it's better to know less.
This is actually empowering in a way because this is, like, return to monkey, you know, and, that's it's the accelerationism has been really worrying to me, lately. But but but actually, in a in a way, it's it's almost doing the opposite. And but but we're still figuring out, like, what is what is the actual career path? We don't know. You know? Like like, I I've used the term senior vibe coder, but I don't think that that's it. I think that there's like, maybe prompt engineer is, like, you know, is a real thing.
[01:05:44] ODELL:
And But it's it's already a real thing in, like, the arts. Yeah. Right?
[01:05:48] Alex Gleason:
But that that's such an easier job than being a senior developer. Being a senior developer sucks because because, like like, you you have no idea what my carpet looks like from pacing back and on it and just, like, the tread of my heels, you know. Like like, this, I've been I I've used my brain so much that that I I get headaches. And, it's been empowering in in, like, the past week or so just building Nostra hub and just just, like, not having to use that part of my brain so hard. And and I I feel like maybe we're actually moving towards the world that we've dreamed of where, like, you know, people people actually don't have to work as hard, like because there has been that promise for decades that we don't have to have the forty hour work week. We can have, like, you know, half of half of that or less.
And and and, like, you know, the rich are taking more and the poor are just getting poorer and stuff. And but may maybe AI can change this. May maybe we can actually do stuff and just not have to work so hard and that that's exciting. That can be empowering.
[01:07:01] ODELL:
But maybe on on that on that note, like, how do you envision stuff like this getting monetized? Like, doesn't it just commoditize a lot of stuff?
[01:07:10] Alex Gleason:
I think that that still you're gonna have people that just don't know anything. And but they but they have they wanna run a business. And they need, like, a consultant to come and help them run their business, someone who does know AI. And and so, and so I think there's still opportunities, like like, you go to the same way that I used to go to, like, pizza shops or whatever on South Street in Philadelphia and be like, hey. Let me build your website for you for freelance. Like, we're seeing people just do that with AI now, and they're just doing and it's easier. And it's more fun, And it's just better for everyone involved.
And, like, it doesn't mean that you have to charge less for it. So
[01:07:52] ODELL:
But doesn't it? If you if you you're it's it's effectively commoditizing that labor. So, like, shouldn't it But I Shouldn't it? I also think there's skill Compress the margins?
[01:08:03] Alex Gleason:
I think I think that that, like, the the there are definitely entire companies that are, gonna be wiped out by this. Mainly, the people who send me spam emails about trying to upgrade my WordPress site.
[01:08:20] ODELL:
But I mean, what about the big tech companies? Like, Jack's business got 12,000 employees in it. Like, what what do they do?
[01:08:29] Alex Gleason:
I I think that it's definitely gonna change. But I think that there's still, like, there's still skill that you can like, the people can attain there and there's still, like, smart smart people are still gonna be able to succeed in life, basically. And so in terms of, like, the really long term bigger picture though, I I don't know if you've read that, like, AI twenty twenty seven thing that these researchers put out that basically said the robots are gonna kill us.
[01:08:58] ODELL:
And Yeah. There's, like, three different paths we can go down and
[01:09:02] Alex Gleason:
Yeah. So so before the robots kills us. Before the robots kill us though, we create utopia. And, and I think that I think we might be headed towards that right now. And if we can just do that and then make sure that the robots don't kill us, then I think that we're gonna be we're gonna be good. We just need c three p o and not Terminator, and then we're good. Because the Androids are gonna be real.
[01:09:26] ODELL:
Yeah. I mean, I talked to someone who's much smarter than me and older and wiser, that I have a lot of respect for. And the way he described it was the humble way of looking at it is right now, if you were to dot plot, like, the different paths for humanity, like, we can go in all these different directions. So, basically, the exact opposite of what that 2027 thing was where they tried to, like, nail down on what was gonna happen because it's more provocative. It'd be the widest possibility set that we've ever had, period, Like in humanity. And and so as a result, really, nobody has any idea how this stuff compounds and what are the knock on effects and where we end up.
But if you're able to adapt and be flexible, that with that with that chaos and uncertainty comes opportunity. There's a lot of opportunity potential there because there's no set path. We don't really know where we go next.
[01:10:28] Alex Gleason:
Exactly. That's the key, I think. It's just the flexibility to change and adapt to this stuff. Because yeah. I think that there's there's still gonna be plenty of opportunities. I I don't think that we're all gonna die in the streets. I think that that we're gonna be okay. But Yeah. I get it. We just we just need to maybe maybe I was thinking about, maybe making some giant Faraday cages for humans to hide in just in case the robot start to launch an all out attack because I think the robots are gonna need a Wi Fi connection, because because, like, the big problem is that you can't really run the model in, like, a inside of an Android. The model is gonna have to run another little server. They're just gonna use Starlink.
Yeah. But if we if we all go in, like, a giant Faraday cage, then it's gonna block all those signals. So they're gonna have to, like, you know, be launching nuclear at us. But but but that that is that's basically, like, you know, another end of the world's possibility anyway. It's just like nuclear war and that and always has been. So, like, you know, is is that is AI a greater risk than the risks that we already had? Or or, like, just an asteroid hitting the earth or anything, you know?
[01:11:42] ODELL:
That twenty twenty seven piece thing was a little bit uncreative because, like, basically, the end result they had was eventually due to competition, both China and The US will plug the AI into the nuke system and then they'll nuke each other. But it's like, okay. Well, the current reality for the last sixty years has been the humans are standing there, and they're gonna nuke each other. So do you don't really change it that much. It's just who to who presses the button or what presses the button. And hopefully, we don't do that, but I would not be surprised if we do do that.
And they're probably already kind of doing that with Palantir and Andrew and stuff. And the incentives are very much aligned with doing that because if your enemy is doing that and they're able to respond much quicker than you because you require human intervention, then, of course, you're gonna plug in your own AI systems into the command and control infrastructure so that Yeah. You can compete. And then we kinda end up in a game theory situation. But humans are resilient, and I'm I'm not overly concerned about it. I'm fascinated by it. I love the idea of uncertainty and and opportunity.
So So I have another question for you. So the the question I gave you was, like, kinda with my ten thirty one hat on in terms of, like, monetization. And I do think there's a lot of opportunity there. And I still, you know, I think as far as successful businesses go, I think the number one fundamental pre AI was who is the founding team and who do they put around them? Like, who are the people? Right? The people are the are the actual value behind the company. And I I don't think that changes in this world. Like, I think, ultimately, like, you know, you're just you're you're making a a a calculated risk on who the people you are that you're backing.
It just changes the toolset they're using. But, now I'm gonna put my OpenSats hat on. You know, me and Gigi have talked about this at length. I don't know how we proceed with OpenSats, and I'm kind of curious as someone I respect deeply. I'm kinda curious on your opinion on how we should think about this stuff at OpenSats because, historically, the main way that we approve grants and and the proof the grants that we put out is, like, massive proof of work in the open source ecosystem. You're, like, submitting you submit your code for review. We're seeing all your commits, whatnot.
Like, how do we evolve in a one shot or not? I mean, we're being provocative with one shot, but Mhmm. In a vibe coded world where maybe in a week or two weeks, people are making apps.
[01:14:26] Alex Gleason:
I mean, if I were you, I would reward good ideas, and I would and I would do smaller, more frequent donations, basically. Like like, someone builds, you know, Vibe Coding Geocache app, like, that's amazing. Here's, like, $5,000. And then if they keep doing it, then, you know, give them more money. And and maybe just, like, look for these people and just and just offer it, basically. That's how I would do it. And and, like, you know, encourage people to to be by coding apps, and and, like, yeah, like, smaller smaller, more frequent payments maybe is how I would consider it.
[01:15:06] ODELL:
What if the person submits, a clearly LLM generated SLOP application?
[01:15:14] Alex Gleason:
Does that just automatically get thrown out? If I mean If you could view it with LLM, like, how do how do we how do we do I mean, I I would I would just click through. I I you can see, like, the level of effort and and the ideas. If if it's Slop, no. I wouldn't find Slop. But but, like, you know, I I would I would reward people for building cool things with AI for sure. And it I I think it doesn't have to be a lot. I I think that, like, a lot of people are gonna keep their day job, and that's okay. Like, we should reward people who are keeping their day job and who are vibe coding awesome stuff on NoScript.
[01:15:47] ODELL:
I see, Pseudo Carlos. I think this is half a joke, but I'm kinda curious on your opinion. He's saying we should just add just replace the board members with AI and just have AI determine the grants. What what are your thoughts on that? Do you think there should still be humans involved in the review process? Probably, there needs to be humans involved in the review process. Because you could just automate it.
[01:16:10] Alex Gleason:
I don't I I don't think I don't think AI is gonna be good at recognizing what's what's good in terms of product like output. I think we need we definitely need humans. You know, I ideally, the humans are also acting as like a wingman, where they're kind of see overlap in what people are doing and be like, I'm gonna put you and you together. And and you've been a great wingman for me, Odell, and I appreciate it. So I think that I think that that's what the community needs more than ever right now is is just, you know, putting the right people together and and and kind of, you know, smaller, more frequent, rewards for things.
[01:16:52] ODELL:
Okay. Awesome. We're pretty aligned. This is a fascinating conversation. I really enjoyed it. Likewise. I I would love to make this more of a regular thing. Are you cool with that? Like, maybe, like, every four or five months or something like that. I would love to do it. Gonna have it's it's gonna evolve quickly. So I'd love to just keep having you on. I mean, you you better get in touch with me, like, next week because things are gonna change so much between now and then that, like, I will I'm down to have okay. We'll have so much to talk about. I I love it. I'm down to have you on more often, actually. As as you'd be down, I'm down. Okay.
Because I think this is the place to watch. Before we wrap, what I asked you as, like, a case study, but, like, do you think it's worth it for us to mirror Elon's tweets to Nasr manually? All of Twitter? No. Just Elon? Just Elon. Yeah. I'm just asking Elon. We because we have Trump. I feel like if I need to follow two people in the world to see, like, how they're affecting my life, those would be the two that would be, like, Trump, like, literally, like, I I see the price of Bitcoin move, and I'm like, oh, I gotta see what Trump said. He must have said something.
[01:18:14] Alex Gleason:
Yeah. It'd be interesting. It'd be one of the few places on the web you can see Elon and Trump in the same feed. And it'd be one of the few where it's signed and hashed at the time of post and can't be deleted.
[01:18:25] ODELL:
Yep. I mean, Elon deleted all a bunch of his tweets. So, like, as an archive mechanism, it's pretty powerful, I think.
[01:18:34] Alex Gleason:
I agree.
[01:18:35] ODELL:
I mean, we're like, Trump's every Trump's Truth Social post should be should be signed and archived. Like, that's what it should be. And if they're not gonna do it, then Yeah. Because he does delete his his post on on Truth Social, and and you can still comment on them on Noister. It's interesting. Exactly. Okay. Awesome. I'm glad we're aligned on that. I'm before I have you on next time, I'm gonna I'll vibe I'll vibe some stuff up with with, gets start dev. And I'll reach out to you if I run into any issues because
[01:19:08] Alex Gleason:
The biggest thing we're working on right now is we're creating a web UI code named Shakespeare. And and this thing is gonna allow you to just go on a web page and start typing prompts, and it'll build your NoStar app. Come out?
[01:19:21] ODELL:
I am What's taking so long? Why don't you just one shot that?
[01:19:26] Alex Gleason:
So so this is gonna be based on NoStar service providers, which is a new spec developed by John Stabb that is supposed to replace the DBM spec. And I'm just I'm just trying to make sure that I this is Discord Discord coded specification, basically. I'm just trying to make sure I have a good thing that people will want to reproduce, and and then, you know, we'll have a sort of a decentralized thing on Noister to where people can build Noister clients. And, you know, I'm I'm racing to try to get it out, like, you know, next week. But Oh, shit.
But, you know, we'll see. It might I might be the We could have that soon. Way there. We'll have it soon for sure.
[01:20:06] ODELL:
So that whole process of using using terminal and and doing it through command line is is about to be unnecessary.
[01:20:15] Alex Gleason:
I hope. But don't wait. Try it out today because it's worth it's worth trying it in the terminal still because it's producing really high quality, you know, one shot implementations.
[01:20:26] ODELL:
So you think what? Do you think there'll be a trade off using the web
[01:20:29] Alex Gleason:
the web UI? Or I think I think it's just that I know that we've got it all figured out on the in the terminal, and we don't have it fully figured out in terms of, like, we're running a service now and especially concerning the money. But that's that's gonna be an interesting thing because, like, the it's gonna go through us, basically. And and other people will be able to run their own too. But there's potentially also, like, a, you know, financial, you know, possibility here.
[01:21:00] ODELL:
Got it. I guess one last question before we leave you because I see, like, a boss commented. What is your thoughts? I mean, I have my own very opinionated thoughts on this. But we were talking about archiving. I mean, you have a lot of history with Mastodon and the Fediverse and the negatives there and also with app protocol and blue sky. What is your opinion, to the people that say, you know, Nostra is actually not useful for archiving data or preserving data because it can't scale because very few people actually run relays.
[01:21:40] Alex Gleason:
I mean, I think that's just wrong. Like, just to to prove a point against that, like, I built a Google Drive relay that just, like, copies your post onto Google Drive. Like like, the fact that it has a signature on it and it's completely portable means you can just set something up to to copy stuff in real time. Like, yeah, it's maybe a problem to query, relays because they might have rate limits or they they might block it. But if you're if you have, like, you know, an application that's open in real time monitoring those for posts, then, like, nothing is gonna stop you from being able to to archive that data forever.
[01:22:17] ODELL:
Yeah. I mean, I, I like to use the exam I mean, now I have a great new example, which is which is Elon's Epstein tweet where he's where he claimed that, Trump is a pedophile, and that's why he's not releasing, the Epstein lists. He deleted that post. If he had posted that on Noster, thousands of people would have saved it because it's signed JSON. You could literally save it on a USB drive if you wanted to. And at any point in the future, you can plug that USB drive in and broadcast it back to the world, and everyone can verify it hasn't been modified.
[01:22:50] Alex Gleason:
That's a great example for why we need to to mirror Elon.
[01:22:54] ODELL:
Exactly. And there was a missed opportunity, but we'll have more opportunities in the future. But I, like, people don't realize that that's that's why Nasr scales so well is because you don't have to run Every relay doesn't have to hold all the information. Running a relay at the base level is actually not that performance intensive. It's very lightweight. So the question becomes so and then so you end up in a situation where you have relays of all different, use cases, and you have people holding nostril events in all different ways. And but, ultimately, it's all interoperable with each other. And at at the at the very like, we could we could walk around with signed paper JSONs if we wanted to. You could put it on a fucking carrier paper. A QR code.
It wouldn't it wouldn't scale that well, but it would work. And it would it would still pass. And I think what technical people have issues with is you don't have this hard availability guarantee. But in practice, it's basically the Streisand effect. Like, if anything is worth keeping, someone and multiple people will keep it. And we see that with the screenshots, people kept you know, there's no protocol for it, but people kept, you know there's probably, like, 20,000 copies, 100,000 copies of unverifiable, which is the problem with it, screenshots of the Epstein post.
And it's because people, you know, thought it was important, and they wanted to save it, and they thought he was gonna delete it. If Kanye joins Noster, every single one of his notes will be saved by a lot of people because he's fucking Kanye.
[01:24:27] Alex Gleason:
Yeah. I I agree wholeheartedly. That's my rant. It was something Vader said I think that, like, the greatest innovation of No, sir, is not the relays. It's the keys. And I I agree with that so much. Because we can Everything is touched. Once the data is signed and it's portable, you can just put it anywhere.
[01:24:43] ODELL:
Yeah. So mine is asking if you can run a relay on the start nine. Yes. You can right now, but it doesn't have clear net support. It should have clear net support any day now. For start nine in general, the biggest issue is that a lot of it relies on Tor, and they're trying to make clear net easier for people. So that should happen soon. But you yes. You can run a relay on a start nine. Alex, this was awesome. Do you have any final thoughts, for the freaks before we wrap?
[01:25:14] Alex Gleason:
Yeah. Like, you know, stick with the vibes.
[01:25:18] ODELL:
Stick with the vibes. I love it. I'll put links to everything we discussed in the show notes. Alex will be joining us again sometime soon. And, Freaks, I gave you we did two dispatches this week because I'm gonna take a little summer break, from dispatch. We will be back on July 2. I'm having, Perveen on, the developer of Cove Wallet. So in the meantime, enjoy your families. I've still be doing rabbit hole recap every week, but just taking a slight break from civil dispatch and catch up on the back catalog. There's, like, four hundred hours of CIL dispatch in the back catalog if you haven't listened to it. Once again, best way to support the show is actually sharing with friends and family. We're available on every platform.
Podcast apps, open up their phone, go to the podcast app, type in still dispatch, press the subscribe button, they'll see it pop up later. And they'll be like, what the hell happened? But thank you all for joining us. Thank you, Alex.
[01:26:29] Alex Gleason:
Thank you, everybody.
[01:26:31] ODELL:
Keep vibing. Vibes are high. Stay on the Stack Sats. Love you all.
Alex's Presentation at the Oslo Freedom Forum
Challenges and Opportunities in Decentralized Platforms
The Role of AI in Decentralized Social Media
Happy Bitcoin Friday
Guest Introduction: Alex Gleason
Truth Social
Challenges of Centralized vs Decentralized Platforms
Bridging Platforms
Limitations and Potential of Mastodon and Bluesky
The Future of AI and Vibe Coding
Empowering Developers with AI
The Impact of AI on Software Development
Building with Getstacks.dev
Impact of AI Models
Monetization and Future of AI Development
Open Source Development in an AI World
Data Preservation Using Nostr