Swarm Community Call — August 2025

[Applause] [Music] You can hear me. Okay, that's it. I've been chosen by the gods to deliver this. Okay, sorry Mina, you're going to have to sit this one out or lucky you, but I'll take it from here. No problem.

Okay, thank you everybody. Couple minutes late and starting technical difficulties with Discord as always. Um, RP Mina, thank you. Thank you for the effort. Um, okay.

So, today, what are we going to do on this community call for Swarm? I'm Callum. I'll be hosting today. I'm sure most of you are familiar with me. Um, for the agenda, what I'm planning on moving through is the core development updates with myself and the B team as well.

Then what I want to do is move to an infocus look at Swarm and our path ahead with basically our partnerships, business development and kind of product and project outlook. And then equally um moving into more of a community talk of web 3 in general. Where's the momentum? Where's the market? what are the policies and where does that put SW form in the mix of everything and is what we're building and the bisdev direction we're moving in sync up with that mission and that narrative and hopefully that's going to be pretty interesting and enlightening for us all definitely have a few good talking points with that being said then what I want to cover is we just had 26 the network and the dashboards that I was looking at myself look better than ever really um price is on its way freezes all-time lows.

Network speeds, upload speeds look great. All in all, really good even distribution of nodes across the network. So, heading towards the B2.7 release, then what we really want to hear from is the B team. And today we have Colin for that.

So, Colin, if you can raise your hand, bring you to stage, and we'll hear what the the latest and greatest is from the B team. Here we go. Okay, Colin, are you there? Can you hear me?

Yeah. Yeah, I can. Can you hear me?

Perfect. Yeah, your mic works as well, at least for me.

All right, off to the races. So, Colin, can you tell us basically how did the 2.6 release go? What results did that give you? And then what are you planning or what are you working on next for the 2.

7? Oh, I I think you you mentioned uh very well uh things look uh very smooth smoother the the we had some improvements regarding uh feeds and uh uh startup times u for also for first time nodes that came into the network and u yeah yeah it's good to good to see that uh uh things are progressing so well and the project is uh is improving uh from release to release. Um okay. So regarding the the status uh the B status for for this uh month before I start with it. Yeah, I'm Khalen.

I'm a big core developer uh and um I'm replacing Nikki for for this month with the status and uh yeah I want to say that uh we've seen strong progress across key areas that are shaping the future of sore network and yeah let's let's get into it into some concrete stuff. So first the B team has been actively working on the was implementation which is a key part of our B in the browser effort. Uh one major milestone here was and is the secure websockets communication specifically autoTTLS. We are now in the final stages with this. We are just having to resolve some last uh reachability issues that to ensure smooth and secure connection for the nodes.

Another challenge uh that we tackled was uh the need for multiple underlay address support. Uh after close collaboration with the research team, a few design iterations, we now have final a final architecture in place and implementation is already underway which is which is pretty awesome. Also uh we also aligned with uh with the research team to on how to handle node certificates which is also an important piece of making uh the browser native no secure and reliable and um yeah that's it for the for the in browser and was implementation uh other stuff that we we did uh this month and yeah even if it was it was a month of holidays August we were pretty busy. So moving on, another important area of work this month has been the pool sync sweep. Uh we've now completed the first steps.

The metrics are implemented. Uh dashboards are live and ready. And yeah, with those insights in place, the team has already started uh work on the protocol improvements themselves. again closely with uh closely coordinated with the research team to ensure every change is grounded in the real needs in the real world needs and yeah this work for both of them will continue through the next month uh we are going to refine validate and iterate both of them beyond these specific features and uh maybe here you you call would have more more uh more more insight sites but uh we are also been looking broadly into how to manage sweeps not just a list of proposals but also how we define sweeps life cycle from idea to evaluation to implementation and beyond. The research team is actively involved in these discussions.

And uh on the more practical and technical side, there has been a really productive conversation particularly around the negative incentives and smart neighborhood management sweeps. Two major proposals that would really love uh to start implement them. We would really love to start implement. And yeah, that's it to to wrap it up. uh BN browser is progressing quite quite nice.

Uh also pulsing improvements are in are under underway now backed by realtime metrics and uh across the back the board we're becoming more intentional about how we manage sweeps and uh evolve the protocol. Uh before I end up I would like to to thank everyone for contributing to swarm me being an engineer I really appreciate you running the notes providing feedback uh uh sorry if we don't get uh quite that fast to to reach to you back and answer we will try to to do better but uh yeah thanks for your involvement and uh yeah I'm glad to to be part of of this project and to see uh what can be built and what can be done with it in the future. Yeah, well said Colin and just to that point to that point um as you say we've been redesigning this web process and that comes with having to make priorities and having to make choices but I think every decision that we're making at the minute and what you see in the releases and then therefore in the network metrics afterwards is that every choice that we're making and thing things that we're including at each release is all metrics driven or we want to impact something that has an actual change that can be measured that then reflects a better service for the network in general. you know, which I think we see kind of time and time again on the last couple of releases. So really what I'm seeing is the network itself becoming more consistent, more robust and better metrics and better speeds every time

I have a conversation like this or at least the conversation is always much more focused or getting more focused with every release all the time. We're kind of zeroing in on on everything which kind of leads me to us coming towards the end of this epic of looking for you know stability improvement speed in speeds and uploads downloads you know pulsync speeds and metrics you name it I think everything is becoming great and consistent across the board which is ultimately the the bedrock of what we're building here as an open source project to then equally include all the sweps both internally and from the external contributors because that is what we're here to do at the end of the day not build what we want and do what we want and that kind of ignore everyone else. So I think you can all see that with the negative incentives everything that we're kind of investigating and proposing inside the research team and from the externals as well um of what to do with stake management you name it. So we are here to completely overhaul things to what it serves the network best. Um both from an engineering perspective, performance perspective and incentives perspective.

You can see all that kind of coming soon and I think that's definitely reflected in the B team's performance over the last couple of months in the last few releases for sure.

Definitely. Yeah.

Well, thank you very much Colin. Um, if there's not anything else, then what I want to do is transition from basically here we are with the last release that's gone well with a great network and we know where we're coming to next with the next release, but what are we going to do with that? What are we building on top of it? What does that allow us to build? Basically, with such a great stable network in a great market, what direction are we going in with it?

What are people building on top of it? What's the functionality that we can actually unlock with what we're actually achieving here? Um, so that being said, and then I'd like to invite Romesh and Danny up onto the stage. You can put your hands up. We'll get you up here.

Okay. Hi guys. Can you hear me? Can you speak? Got your microphones?

Yeah. Can you guys hear me?

Yeah, I can hear your mesh. Uh, I think yours is not on or permitted. But anyway, so while Danny's figuring that out, Romesh, basically what we just talked about is the development on Swarm is just accelerating and accelerating. We're getting past the point of stability and more starting to excel in performance and kind of all metrics here to look at the bigger picture and kind of look at the ongoing projects and what we're building on top of. Then give us a sense of where we're going with the products that are being developed on top of Swarm and any of the specific kind of business or partnership directions that we're moving in and and why or what you think about that.

Sure. So um uh thanks for having me. So one of the one of the key areas that we're focusing on currently is accessibility, ease of use and bringing more stability to the network. So uh building out like services on top of the protocol and targeting key verticals where swarm uh has a particularly unique advantage. So uh in in that sense across all three we and the developer community have made quite a bit of progress uh especially in the recent months.

So uh I'm going to start with like accessibility and ease of use first. So with the latest 2.6 release we have seen massive increase in terms of performance. Uh we have reduced the sync time which used to be quite a bit of a waiting period. um especially for new node operators and uh so through a series of optimizations uh in peer discovery and data synchronization we have managed to bring that down by almost 80% in our test environments.

So I think like in the real world we will see um how well that is also turning out. So this makes joining the new networks faster as when you spin up new b nodes and uh the network is uh uh far more reliable than ever. So effectively lowering the barrier of entry and uh so the idea is to bring in more users and uh more node runners as well. So so those are the the improvements that have already been there. So there also been improvements in postage stamp syncing.

So to reduce the bootstrapping time there as well and uh as far and there is actually even more optimizations coming to the postage I'm thinking that's a work in progress so we are going to see that in the next releases as well. Yeah, definitely. And look, accessibility is probably my favorite topic at the minute anyway when it comes to swarm. And equally, all those again being metrics driven, you know, the percentages that we're talking about and that just goes to show with what we're actually caring about. You know, you measure what you care about and I think that's exactly where our heads are at at the minute.

That is, you know, one aspect of, let's say, as a node operator as you're trying to get set up on the network for a node. And all of these startup times, syncing times, kind of the fundamental processes and metrics of the network are all fantastic. That's for, you know, the core infrastructure users or kind of the network participants. But then even if we go one layer above that, we're talking about kind of network usage. And from there, what we're interested in is how easily can you come in and maybe not just as a node operator, but just as a user that wants to upload something, download something.

And you know, we've alluded to that already with kind of Colin already saying that they've got everything set in place to build towards this in browser client. You know, Danny, if he was here as well and talking, he would be kind of emphasizing and yourself with the usability of let's say Bports or, you know, aka the the multi-chain app. So even on the network level, the performance is great. On the startup times and the running times of everything, all the network, all the the metrics are great. And then even above that, the ease of use of the product suite that we kind of have in our hands that is kind of getting ready to deliver over the next weeks and months kind of takes that to a whole different level that form hasn't had before, which myself personally very very interested in.

Right. Right. So um that that's the other thing, right? So there the couple of uh parallel things that are also being worked on from like a more technical side is that there is this uh brand new file manager that we are building on top of this one desktop app. Um there Oh hi uh Danny is here.

I I think his speaker or mic okay so um yeah I'll continue and uh so yeah and we're planning to uh integrate act uh also onto the file manager so that it will make it easier to sync local files with swarm so those are desk around the swarm desktop and improving uh branching utility to the swarm desktop uh app. Uh on the on the other front, there is also a new um a new work being done on bringing global pinning finally to swarm. It's something that I've been waiting for uh more than a year now. Uh so I'm actually super happy that uh global uh work on global pinninging is being done to bring that. So uh just to to give an idea of like why uh global pinning I think global pinning is critical for swarm.

Um so right now we can already pin data to the node but you cannot uh an another node in the network cannot access it. So the pinned data is not directly served um by the network or accessible to the network unless you push the pinned data to the disk. Uh glo what global pinning will bring is that um you will be able to pin content directly to your local node and serve it to the other nodes in the network. So in a way in a way it actually brings back the decentralized culture of self-hosting that we used to have back in the early days of the internet uh long before the the whole centralized servers and Amazon and all the cloud things came into the picture. So I I see this as like a pretty exciting new feature that is coming and also because um global ping will bring a almost free tier to the to the network where as long as you are self-hosting you don't you're not exactly paying for storage yet.

So you can start building an app and serving it and locally pin it uh pin your data and eventually as the demand comes uh or or the as the demand grows then you might want to upgrade it and store it to the disk on swarm where you get all the performance benefits on top. So so yeah global training will in that sense bring uh sort of a IPFS feature parity onto swarm. So I'm pretty excited about new features.

Definitely. And like kind of subtly, I think that's where things are going, right? Which I guess it's just the the UX side or the the feature and services side of Swarm is what's really coming on here. So if we have global pinning, you know, we have great performance, you can run locally, you can build out basically any app infrastructure for your use case kind of in a permissionless way, which you know, completely deviate from any of the other kind of decentralized storage protocols or anything remotely similar. Um when it comes to then services that are being built on top of mesh then what is happening there basically with these features and what we have available who's building what if it's you know bport if it's swarmmy basically what are you looking at that you're saying that's exactly the right use of this feature

right right exactly exactly know so yeah I I I would like to see global pinning come to to this one work. Um, uh, yeah, maybe I I can't wait. I can't wait for this. And

then literally, so even building on top of that, we have all these different things put in place, you know, we have then this fleet of services that can get built on top with everything that we've delivered and the performance that we have. We have this great opportunity to make pretty much any application that we want. And you can see things like for the likes of Swarmmy that allows users to you know pay for storage with a credit card um it's already working with kind of fantastic partners like snapshot or you know another initiative that we have which we've talked about you know you may have heard it as the you know multi-chain app renamed um lovingly to Bport um which kind of takes this whole network to a very different level and something that's incredibly easy to use and then just like Colin said before this Then what we also have kind of coming as a product is the you know at least the demo initially of the in browser client. So to you know square the circle here what we really have is a hyper performant network way for anybody to come along and use it and upload to it and a very easy way to be able to come along and access and download whatever has been uploaded to the network as well. all the while not having to spin up your own node or really do any kind of complicated payment processes by postage stamps or you know everything is abstracted to to the background I think that's just incredibly exciting for what can be built and what is being built already you know even if it's in these key verticals that you mentioned Romesh like you know between what's happened with you know AI and multimedia and kind of what I hear with the MC developments so if you can maybe talk about that for a few minutes I'd be interested to hear the developments Right.

Right. So before uh we go to NCP. So there we also uh have um more improvements happening on on Swami which is uh which is the first um service provider on uh on uh for the swam network. So if you don't want to run your own bnode then you can actually start using Swarmi. Um so so the there there's that track as well.

Then uh regarding the Dport that you mentioned again uh there is it's still uh making a great progress and uh the core infrastructure is ready uh on the Bport and we are doing some heavy testing at the moment. We are focused on scaling it adding more nodes to the network and deepening liquidity for more token pairs so that uh Vport will be um widely accessible and usable. Um uh once that is done we also working on a library and a front end widget so that anyone can integrate this um which makes buying uh say from any chain or any network pretty easy and like yeah like drop integrations possible eventually this functionality will also be integrated into the new file manager that's being worked on. So that's about the Bport. Uh to come uh to the last one which is the MCP and um uh targeting the the whole area of uh the domain of a AI and multimedia.

Uh what we are oh actually the the solar punk team have been working quite um quite intensely on this and so doing some good work there. Um so after a successful proof of concept they are now building a minimum viable product and the goal is to create a seamless integration position swarm as a default decentralized data streaming layer for for AI for video. Um and uh the MCP uh the MCP MVP is is initially is just the starting uh phase of that. So it's all about making swarm the easiest and most secure choice for developers. Oh, and while we on the topic of streaming, um I think uh Solar Punk's team is also working on a demo that Oh, actually they they working on a previous demo that already shows that live highquality video streaming is already possible using Swarm's native primitives and I can't wait to see that also come to Exactly.

And well, I'll give Danny a last chance to see if his mic works. If he wants to come in and say anything about that.

Yeah, I think that Thank you. Can you hear me now?

I can hear you now. Yeah, go for it.

I really don't know what was the issue. Uh I think I think Ramsh just actually went through uh on on everything that we plan to uh to share with you during during this call. um as as as he mentioned uh the solar pan team is working on uh on this uh on this MCB protocol uh MVP and I think that's going to uh let us uh tap into uh this AI uh AI market which is uh which is definitely booming. One of the one of the key segment that we see there as having a having a unique opportunity for sparm is agentto aagent communication and the ecosystems around that and there are many small AI companies uh trying to find a way to cooperate with each other what their struggle is uh that if they want to share data how they are going to do that and what what's going to be the secure layer to share data and that's where swarm can come into the place so uh so uh with CT uh we are absolutely brilliantly positioned uh to give a solution for these agentto agent communication protocols. We are in discussion with many key players in the market uh and I'm sure that you will be hearing uh quite a lot about that in the in the future.

Yeah. And to then close this this topic off really or at least the thought that's in my head is where is the market going for crypto and development and AI in general and then are we building the right things and measuring the right things here at Swarm and is that what we're reflecting here on this community call. So I'd say what we have is a hyper performant network that's incredibly extensible and then building out all of these easy to use features product suites and connections with MCP to be able to extend into where crypto and tech in general and where the world is going basically from just crypto in general storage in general bandwidth video multimedia AI privacy first you know I'd say the beautiful thing about Swarm is that it can extend end to all of these use cases even all within one product and which is what I'm really waiting on in some way shape or form but that's what I'm excited by all everything we've just heard for from the network performance to the releases to the products and then what's ahead when I weigh that up with where the kind of the world is or where the market is at the minute or where the the technology is I see us moving right along with it um but more to that point the man of the hour um I will move on to basically our next part which is the community talk which is web 3 momentum and crypto policy. So we're going to find out kind of exactly where the market is, what's moving it and what are the trends, what what can we expect over the next lot of months or years even with uh both the tech and the policies around it. So with that I'll invite Lou to the stage.

He's got his sign up already. If I can get light. There we go. You're on stage. Can you hear us?

Okay.

All right. Hey, I Callum, I can hear you just fine. You hear me?

Perfect. Thanks. Yep. Of course can. Yeah.

Thanks for joining us. Um, look, how to introduce you. Basically, you've done so much and too many interesting things to mention, but basically what I should say is this is Lucerner. Basically, the the founder of Crypto Mondays Collective. Uh, basically the largest crypto meetup group on the face of the earth, I would say at this point.

Uh, recently coming to Ireland as well, which is what I'm quite keen about. So, please hear that. But Lou, why don't you kind of give everyone a brief introduction and then I'll set the scene. You're muted, by the way, Lou, just in case you didn't know.

Sorry about that. As Callum As Callum said, um,

you're muted.

Hey, Callum, I'm sorry. I'll be back. Just one sec.

Not a bother. Not a bother. There should be a button where I can play the Discord music while we wait. [Music] [Applause] Hey, sorry about that. So, no worries.

I was saying so uh so my quick background I was a Wall Street analyst for the first third of my career. I started two uh crypto companies. Uh the first was uh the top level domain.tv which we bought from the island nation of Duvalu for $50 million.99 and the second was called bulk.

the largest social network in the world before MySpace. And you know, I spent a lot of time thinking about community. And because I spent so much time thinking about community, um you know, when when I saw the uh crypto light on June 29th, 2017, what I saw was for the first time in history, we have a tool set to solve for uh the community instead of the man in the middle. And I thought that was going to be the biggest thing in the history of mankind. and I've been 247 uh crypto for eight years now.

Yeah. Illustrious career to say the least. Um so that's why you're qualified basically to have this this conversation with me on this uh community call for web 3 momentum and the crypto policies. So just as an overview, we're building all of this on Swarm. We've have all these features.

We have all this technical development. we have this explosion of of AI and you're equally very deep with all of the AI updates as well. I probably don't know anyone that that is more kind of in the mix of where AI and crypto intersect than you. Um, where do you feel the momentum is with web 3 at the minute as well as the market conditions and basically this overhaul in SEC policy becoming extremely favorable and probably extremely privacy focused which is interesting especially given the the kind of counterpoint or counterargument to where EU has definitely changed their their policies or their outlook as well. So, just wondering web 3 in general, what's your what's your feeling on market narrative, sentiment, and policy for where we are and where we're going to head to?

You're muted, by the way, Louis as well, just in case you didn't know.

Well, sorry about that. Um, you know, obviously I'm I'm very bullish long term and as I mentioned, I used to be a securities analyst and so, um, you know, it's it's hard to, you know, really kind of appreciate exactly the moment that you're at because it changes moment to moment. Uh, you know, since I was a Wall Street analyst, I put out uh two price targets. One was on Facebook, which I saw in the early 2010 years before it went public. And the second is I have a 10-year million- dollar price target on Bitcoin that I put out in January of 20 U 21.

And so and we're tracking very well against that. And so it's a long way of saying, you know, I'm I'm long-term bullish. You know, shortterm, you know, it's like the weather. You know, nobody knows what the weather's going to be in a month. There are just too many variables and there are just, you know, so many things daytoday that that impact the price and change.

But, you know, all of the reasons that most of us got into crypto, you know, are are becoming, you know, more important every single day. And I feel more confident about the the 10-year million-doll price target. You know, obviously, there's been a sea change um in terms of policy, you know, in the United States. And um you know, that's had a certainly a positive impact on price. But for me, you know, I'm not here because, you know, the government is is going to like us.

I don't think the government is going to like us. Full stop. You know, I I don't spend a whole lot of time on on government and policies. I really try and keep my mind share on government spent on two very simple facts. Fact number one, the government is not our friend.

And fact number two, tomorrow it's going to be diff tomorrow it's going to be worse.

Always the optimist. Um Yeah, I mean I I feel like I'm an optimist. I do. I I totally feel like I'm an optimist. You know, that's why I'm out here and that's why I believe, you know, that uh you know, that we've got a chance to win this fight.

Well, that's it. I think if you can't beat him, join him. I think that's exactly what we're seeing at this specific minute in time. Um, you know, even very recently, we've got basically the Republicans that are highly involved in the SEC basically coming out and quoting the, you know, cippher punk manifesto for, you know, a heavily crypto.

Yeah. That's because, you know, that's how they can scam us and make money off. It's not how they can help us.

Absolutely. Um, but to give a very very um obscure, you know, Irish colloquialism, you know, what's good for the goose is good for the gander. And you know if that helps us in the meantime at this minute in time with the market.

I completely agree. Yes.

Yeah.

I completely agree with you. Just just you know we can't delude ourselves that in fact they are our friends and they're here to help us and they're going to help us. They're they're not going to help us

unless it's by happen chance because it helps them.

Yeah.

Yeah. So the take the takeaway from this is that if if the government even the government is talking about privacy whether we believe them or not we are on the wrong on the right track keeping privacy as the main focus. So that's optimistic takeaway that I see.

Well I completely agree

regardless of what the Yeah. Regardless of what the government does it's in our best interest to embrace privacy.

Exactly. Yeah, which you know and this kind of goes back to you know the section before this is you know are we building and are our beliefs aligned with the market and where we're going long term and I would say absolutely with this measure you know everyone will adapt policies will open up and then they'll close again you know uh administrations can change every four years it will do uh and people will change their stance on policies and privacy etc but the difference here is that that may be very well true for centralized services, anything like that. You know, the EU is trying to consider or at least considering banning, you know, encrypted messaging, you know, things like that, just active monitoring. And that comes down to the the principle of, you know, is do you can I write code based on the law or is code the law? And at least with Swarm, we can say that well, whatever we do, it follows the code.

We're privacy first. We're permissionless. Nothing can touch it. That is kind of the guiding principle of crypto and privacy and what we're here to to work on and deal with anyway. And I think that is where the market's heading though to your point.

You know, for how long we'll see, but I don't think that they can slow down the technology and I think that's what's reflected in the current market prices.

Yeah, I I I agree with everything you said. I for me I don't think the government though is really um you know are you know our I mean well while while I I do think they are our enemy you know luckily they're you know mostly really incompetent um yeah and you know I I think our challenge is is not the government but I think the biggest threat is uh what I've been calling for uh eight and a half years now fam which is Facebook Apple Microsoft, Google, and Amazon and the fact that, you know, we're now, you know, living in in a world of a few tech monopolies. And, you know, they they also do not want this to be successful. And so, you know, they're the ones who are really going to try and capture, you know, all of the value from this.

Exactly. Look, we don't want them to be just continue to monetize every ounce of data, you know, across the the face of the earth. And then to take it one step further for for privacy as well, you know, the your personal private messages or your private data is not your own. It can be subpoenaed. It can be obtained again by the government, by lawmakers very very easily when it's not theirs and shouldn't be readily accessible.

You know, that's where private messaging, encryption, swarm, any service like that is just incredibly important and kind of constitutional as a private citizen that you kind of have the right to privacy at some degree and that has to be reflected in the technology that you build that you back and that we use as a society. You know, it's very web two way of thinking for everything that happened beforehand of how they siloed every ounce of data that they could get their hands on. But again, it's brave new world. I think there will be differences.

Yeah. I mean, yeah. Go ahead.

No, I was just going to say that, you know, obviously we're moving now, you know, as you've been discussing, you know, into a world of AI and that's going to be such a a dominant technology and one that, you know, appears to be driven by data and those guys have so much data. Um, you know, and so, you know, I think it's also, you know, incredibly important for us to, you know, really focus on, you know, the needs of decentralized AI and and have that as an alternative for the people who want it.

Exactly. So that that was going to be my my next and most important question is with the future of web 3 and the ecosystem where we're kind of coming to the intersection of where does AI decentralized AI you know decentralized protocols and storage kind of all come into place you know what does that future look like you know as opposed to just running you know uh llama or a different kind of you know closed source model where they've just harvested your kind of proprietary data instead it needs there needs to be open source models that we can kind understand Korea and add to and trust. Um where do you see AI decentralized AI and the various protocols that kind of enable it like swarm kind of where does that come into play in the next few years? Yeah, I mean I think you know certainly a everybody knows that that AI is coming um you know and I think you the products that Swarm is building you know to enable that um you know is you know certainly they're going where you know we hope the market goes in terms of uh you know having at least an option for for decentralized AI. Um, you know, it's it's my belief that, you know, the main way most of us are going to broadly interact with AI is through AI agents, right?

And so, you know, that's why, you know, we're we're both a part of the decentralized AI agent alliance, right? And and setting the rules, you know, or the standards for how that works. So it's able to scale as as you know we can scale these decentralized AIs cuz I think that's you know I think decentralized AI you know AI in general I think of it like electricity you know one day there wasn't electricity and then there was you know electricity the next day and the first people who had it had a very big advantage but you know very quickly everybody got electricity and today you know very you know electricity powers yeah 100% of everything but very few people have an electricity advantage Um, you know, hopefully that's where, you know, we're going to go with AI. We will have decentralized AI alternatives that will be just as good as as the centralized AI choices.

Yeah, absolutely. I definitely agree with that and that that is kind of fundamentally important for maybe a couple of reasons which I don't know if maybe just based on all the projects that you come in contact with that are let's say a mix of like decentralized and obviously we've all used and know the kind of centralized models as well but you know they can be censored not give certain replies hallucinate you know based on the data that they're given and definitely be driven to respond in a certain kind of way as opposed to let's what a decentralized AI would not do and is actually free to operate whatever way they like or you can fork it and do that yourself. So, I'm just wondering, have you seen any projects that are doing that really well for decentralized AI agents that kind of might verge into the world of of kind of crypto and using decentralized permissionless protocols as well as opposed to use the

large models?

I I I I think we're incredibly early today. Um you know there was a lot of excitement around you know a AI agents you know maybe about a year ago um year and a half ago now and you know there was a lot of enthusiasm around it um and you know uh uh it's the the bottom line is stuff takes time and so you know I I think you know you saw kind of some of the early leaders like uh you know Eliza um you know come out and and have some you really interesting like really simple bots. Um but yeah on the AI side you know what what we found is is that in fact um you know e even simple things like ordering Door Dash for example um you know I heard this company speak you know who'd been you know the guy from Google DeepMind you know built a team and you know they worked for 6 months trying to have an AI agent that would be able to handle your Door Dash and you know they were only able to get the the uh order right like 60% % of the time, which obviously is so far from, you know, the 99 plus that it needs to be. Um, you know, and it shows you where we are. We're just we're just very very early.

You actually I heard the guy from Eliza was on a panel recently and you know, he asked, you know, what's what's the main problem holding back faster growth of of AI agents? And he said, well, I think the main problem is they don't work. And that's really the truth where we are today. It's very very early but um you know we have a lot of the smartest people I think in the world you know are working at the intersection of you know AI and web 3 and um you know I think we're going to solve our our problems.

It it definitely I think just in that vein for most of these problems it comes down to what you said before is just data. There needs to be a lot of data that we need to get our hands on to be able to sufficiently train these models, you know, or run it on synthetic data. You need somewhere to put that. You need to be able to train these decentralized models and agents at a better, faster rate that's more productive than say what the large scale private models are doing. Um, and I think that's where, you know, decentralization and open source projects maybe like Eliza and like Sporm kind of all start to kind of fit together, you know, fit like a glove where we can do the right things.

and we kind of open source that put as much data in as we need to and we can build and iterate whatever we like in a permissionless way and everybody can avail of that not just privately. Um, so I think you're right. It's going to take a long time. It's going to take a lot of people and a lot of protocols coming together um to have it work properly and in an open source way ideally. Um, which would be nice.

But I think you're right. It is going to take an awful awful long time and it's rarely the first um out the gate that kind of, you know, wins everything in the end as well. And it's probably not a winner take all market. Again, the world we're going to end up in is very agentic where there's, you know, swarms of agents that do many different things and interact together across, you know, a numerous different protocols and, you know, you have a fleet of them in in web 3. And luckily, just happy that Swarm performs a very key core function and can integrate with, you know, all of that either if it's with um, you know, our MCP efforts that, you know, Danny just talked about earlier as well.

It's just a great start. Yeah, I mean I I really believe that, you know, Swarm's deep embracing of AI is, you know, ultimately going to serve, you know, the the entire community really really well.

Absolutely. Like likewise. But with that, that brings us towards the end of our community call. So know we've got 10 minutes left to go. So, I thought I'd just open up to quick Q&A if there's any questions in the chat for anybody on stage whatsoever.

And easy ones, please. And we'll be happy to answer them. Give a few minutes to see if anything comes through. Um, I can uh if we are while we wait for questions, I can add something. So I completely agree with what Lou mentioned about um being early and we are actually super early um in both in terms of recognizing the value of privacy which is why you have um Facebook and all these kind of centralized data collectors uh extracting so much value.

uh now you see the same thing with AI and I think what we are going to see in down this line is like when now that even government is quoting from the cypher punk um manifesto and like talking about privacy uh you know that okay this is privacy is going to uh become a far more valuable asset down the line not to trust the government but at least we know that uh there is a gradual awareness on the importance of privacy and I see privacy becoming uh more valuable assets down the line and so that's when I think we will see the real adoption that we all want but we are getting there if I may add to to something to this as uh as Lou mentioned uh he used the analogy of of the of the electricity as a commodity uh and and if you use that analogy nowadays you are plugging yourself into the into the grid and you're getting electricity in the and you don't even know where the power plant is and I think that's that's part of the future what AI uh what AI will see is like now we are talking about these large scale big models dominating the market but there are like many many different uh different uh models being built and and and I'm I'm sure that they are going to be specialized in certain areas and uh the value that the actual customer is going to get at the end of the day it's going to be like like electricity they will not know where the data came from and and and with what interactions uh different AI agents uh made just to just to make sure that that he's getting the right answer and this is where where swarm with its uh with its uh security layer is is is very well positioned. So we if if we can only be uh the the transition the the transactional layers behind be between these agents that's already uh that's already something which is uh which is uh which is great to see.

Yeah, well said. Thank you very much Danny and thank you very very much everyone else for coming on stage and talking and thanks everybody for attending and listening to this month's community call. So no questions this month but then we will see you again on the next community call. I believe that's the 25th of September. So see you all then and thank you very much.

Bye.

Thanks everyone.