Protecting Digital Public Spaces in the age of the AI Botnet Swarm

Transcript

All right. Well, sorry for the metallic for not using DWeb, um, but it was only signing with Privy. And ever since they were acquired by Stripe, I haven't been too keen to sign with Privy even with my own PK. Um, but alas. So, thank you guys for coming to this talk. Uh, it surely is my pleasure to give it. My name is Taylor. I go by Polysaturkiss online. And I suppose a fun fact about me is I was born in the same city as this guy, Marshall McLuhan. Uh, he was definitely before our time. If you search him up on YouTube, if you're graced enough to uh get search results from YouTube, it's all in black and white. Uh, but he was a prominent media theorist. And he coined the phrase the medium is the message uh

because the message shapes and forms uh the search and forms, sorry, of human association. And I think it's just a really relevant quote to lead this talk with called protecting digital public spaces in the age of the AI botnet swarm. Now, I thought I'd divide this talk into three sections. I like to start with a bit of a historical framing uh to help justify why we need to re-posture our online services just to have the same amount of feature parity that we used to enjoy in the age of post robots.txt. I then like to derive for you guys the economics case for the world computer.

I'd hope to be able to convince you that by judoing the internally online and escalating shouting match that we can leverage it against itself for the industry and public benefit. I then I'd like to cast the remaining time to explore these architectures as dapp developers to how you can bring it into much more into Twitters, into YouTubes, into any sort of comparable service. So, by the time it reaches maybe 2030, we have a Veracity Net, essentially. So, to begin, there was a dream that was the World Wide Web. And and to preface this, this is very different than the dream of the internet. Probably a bit hard to see here, but this was the original well,

semi-original. This was ARPANET brought up by DARPA. Uh just down a little bit there, you can see Eglin Air Force Base. We'll talk about them a little bit more in our search for creating savory scapegoats for a conversation such as this. Um but the internet to the World Wide Web was the academic portion on top of I'd say the deep state portion that spoke to the universal access to the world's information. It's almost an oxymoron to say we have many universities. A universe is by definition one, but it's the first time that we could create a universe of universities with complex data types.

We take for granted the hypertext part of the internet for today, but being able to click on an A graph was revolutionary for the time. Now, he built an academic file sharing protocol, not selling. And that will be very important for us later in the talk. If it was the metaphysical postulate of the space age to go where no man has gone before, then let's make it the metaphysical postulate of the information age that you affect, you change the world by what you choose to pay attention to. And that's the secret power that we're going to leverage here today.

And as an overarching question I like to leave the audience is what does that mean if we only really can physiologically tune in to bot spam? Now, maybe it was Tim Berners-Lee himself that committed the original sin by adding the 402 status code uh for payment required. Uh but as engineers and developers, we can really appreciate that we take an academic file sharing protocol and try to do something with it after, something's going to break. So, we can see here in the fallout of monetization in the decades that ensued, it started to look like digital oil.

Advertisers were saying, "Hey, can we just have your user information? We can just better Jones effect and pay grassroots style campaigns." They're like, "Sure, okay, here's our user data." It started to look like digital gold. And that's when we really started to see HTTP and TCP over IP really being exploited in essence against itself uh for for profit. And so, I would say by around 2015, 2016, this was clearly a digital gold rush. No one was well armed to like to wrangle the behemoth. And we saw a paranational struggle for narrative dominance ensue. After we saw things like Cambridge Analytica vacuum up vast portions of the net, gather maybe 200 to 3,000 weekly correlating

behavioral indicators on do you go team A? Do you go team B? And then doing advanced algorithmic things like max cuts to separate people into target and persuasion groups to be behaviorally nudged. Now, if the users are most affected, um if you wanted to run a Youngian word association test in psychology, you'd have to go to your board, you'd have to ask for permission. Uh you don't have to do any of those things online. Um so, users were most affected, but in second place was the sysadmins and the moderators of these services who are now subjected to nothing less than state-level espionage.

So, you just showed up to create a website, and now you're playing with nation-states. Um And so, we can see here, I'd say I I'm always picking on Egglin here uh for unsavory scapegoat. But after Smith-Mundt Act passed, they became the most astroturfing uh Air Force base on Reddit. So, if you're going to Reddit in 2012, you're probably talking with some guy in a military suit. Um that being the Cambridge Analytica logo, if you guys remember that. So, fast forward to today, maybe 2025 to 2026, we're in the corporate capture end game. Uh nothing is really stopping Big Tech. It's they're calling them the new oligarchs of the age. It's why make a better product if you can just lobby

these corporate lobby these governments and then enforce the use force the users to hand over the data. It's a lot simpler of a business model. Uh you [snorts] can see it such as the Article 13 debate, I think, here in the EU. Where you saw Google just flooding uh the EU Parliament. You saw anti-Google flooding the EU Parliament and government just being, "Okay, well, we're under a bunch of spam." Uh a couple years later, we saw the DSMA DMA briefing. We saw all the Googles um civil lobbying, forming a bunch of grassroots campaigns saying, "Okay, we're totally not Google right now, but like can you wage Google's agenda?"

Um and uh it came off a Reddit post. Please do your own research in terms of the veracity, but one user traced 2 billion of this style of civil lobbying in terms of the push for biometric at verification and OS identity. And so, while it looks like this lockstep push uh from maybe this kind of like overarching like shadow, there is a lot to be said about just like the raw money incentives in this lockstep um fencing off of the net. Um so, there's no so shortage of people in Web 2 looking for solutions to these problems. Um I would attest that they're all Faustian in a sense is that you give a person one thing, um but they'll cost double of another. If you sign up your

service for Cloudflare, it's like, "Great, you're protected from bots." Uh but now you're giving all the data through an SSL and MITM. Uh Cloudflare to agree with well know it performs SSL termination. So, it'll take the S off of your HTTP, determine if you're a bot, re-encrypt your traffic, and then give it back. And they're now a Palantir partner. So, in terms of to fight, deny, defend oppose embrace extend extinguish, it is definitely a very, very, very black hole. So, in the areas where they didn't go down the Cloudflare route, maybe it was more like Twitter, where it's okay, maybe we'll just start combing in like credit cards. We saw the portioning of the net kind of being

restricted off into speech is not reach, and this is very critical towards our business case that we'll build here today. Because what they found is that advertisers just don't want to be associated with certain information. Users naturally want to corner off into different places, and it's a lot easier just to portion off piecemeal pieces of individuals' identities at different points of time to make a more amicable social net. And so in terms of solution spaces, they're all Faucian for Web 2, but they're just non-solutions for the public space. Like already, like just by having Cloudflare cookies in these public forums, all perceptions of anonymity is gone.

You try to You try to defend them against the bots. There's just not enough money there really to like tackle it at all. And so even though you have your signing these CAPTCHAs, the bots are flooding everywhere. So, something needs to be changed if you want to move forward. So, if the internet's not dead already, then it's certainly dying. Homage to that dead internet theory that made its way around. And it's not just the algorithmic content, but it's also how we're served that content. It used to be that we used to spend months on a certain like certain gate of something, we'd spend months on it. Now, we maybe spend days. And something that would be totally contrite, now we spend years.

It's a total like reframing and recontextualizing what information is and how it matters. And I don't know if anyone can really like feel that too. It's just like there's is a drastic this tyranny of the present just marching on into the next thing into the next thing um in sort of like this hypnotic inducing algorithmic design. And so maybe we can move a bit forward from the dead internet theory at this point as dApp developers and maybe we can start talking about live internet theory that be like a like outside of this doom scroll of like the big five tech companies that there are actual real people interacting with each other each other like benefiting each other's lives. And we can point to things like the Digg shutdown that happened maybe a couple months ago where

they tried to reboot this old social network but it was immediately flooded by bots. And so and if anyone's maybe arguing most people would be sympathetic because maybe like OpenAI closed AI or Anthropic they're not doing much effort in terms of like hiding the AI-ness of their things and I'm pretty appreciative of that. But now I think everyone is really like tuning into the AI mannerisms or the slot mannerisms. So no one I think at this point is saying dead internet is a conspiracy. And so if I wrap all of this up into kind of like a moniker, I like to call it the bots swarm gulags where we're kind of pied piper-in under this illusion of a service and really in the

end we're just there to train their bots and be trained by their botnets. And so if I can ideate kind of that we're living in this bots swarm gulag, what can we do about it? And from here I like to build the economics case for the world computer. And so I'd like to first preface that big tech won't solve it itself. If big tech were to solve it, it would have done it 15 years ago. We find that no there's no individual interest that really wants to solve it.

The adtech as long as they're getting advertisers to sign up, it's okay. The social media just looks like it's engagement and like why would the people faking the engagement be kind of honest about it? Um and so this is a good talk good place to think about kind of like an information iceberg. Um so we can be really cognizant about the corporate actions. It's really good place It's a really good place to take concrete steps. Uh but we should be cognizant about the what's lying beneath the surface. Um and if anyone's aware of terms like full spectrum dominance or fifth generation warfare, then they'd be pretty sympathetic to the idea that they will spend a fortuitous amount of money to win the war for your mind.

And so I'll give a little quote. Nobody said this, but perhaps it's apt enough. Is that give me the corner on the world's market for culture, give me the monopoly on the world's market for culture, and I care not which market you corner in return. If you have a monopoly on oil, I will use the information apparatus so no one wants to buy an oil car. I could have a bunch of useless pile of rocks and I could convince people that it should be the forefront ritual in any sort of marriage situation. Uh if there's a supply glut for coffee, I could say that coffee is good for you.

If there's a supply shortage of coffee, I could say it's coffee is bad. And so maybe it really is a paved paradise paved paradise situation. In the early '90s, we had this like everyone could just have speech is equal to reach. Now you could reach the other end of the world really easily, but that's gone away. And I'd say it's going to go away to a point where we might as well start talking about a new digital dark age where we have about the same amount of connectivity village to village as we did in the Middle Ages, but it's just abstracted behind a lot of bot spam so we feel like we're talking to actual people.

And so when we start talking about our solution space to all of this, I just like to highlight that it's rolling downhill. Every single time that big tech decides to enshittify on its captive user base, solutions that can weather the stand of time look by comparison a lot more attractive. So if you if you want to try to create like a story amongst people, kind of like what's going on in the internet right now, I like to call a story about Dante's swarmers. And you can see how each level below kind of informs the level above. On the top level, you have the engagement farms. These are the people proverbially dirtying their hands on behalf of the party, running the fake bots, buying those huge warehouses you might see like on those like Twitter feeds where they

just have racks and racks of phones creating fake engagement. Below that, you have the PR firms which rent out these firms. So if you're like your billionaire that's committed a bunch of heinous crimes, you want to go and ask them to do Wikipedia brigading for example. Corporations which of course would leverage the PR firms, they're looking at a different flat metric. They're define defend oppose, embrace extend extinguish. They'll do things like grassroots cutouts and they'll look for things, "Hey, let's go and bot swarm a government." Below that, we have the NGOs. We'll find corporations making conglomerates below the surface and they'll want to say, "Okay, we need to wage our strategy amongst up several jurisdictions." This is when the war chest gets a lot bigger and you start seeing those stories about

just massive data vacuuming, sorting people into groups called persuasion targets and then feeding them AB uh Youngian-style uh behavioral nudge testing. Below that, we have the public private partnerships. We have like the Cloudflares and the Palantirs and this is when it starts breaking between kind of that iceberg beneath it below and beneath above and below to where it stops becoming more so about monetary interests and more starts becoming about geopolitical interests. And so below that we have the states stock the stock puppets for example spreading pro-country propaganda. And of course, you can always picture that there's going to be some sort of like meta system. You can picture Dune some sort of like guy maybe in Antarctica

pitting a bunch of bots forms together for a Hegelian reaction problem reaction solution idea. And so I'm hesitant to put numbers on this because it's not the idea. The idea is having good information. But I do want to try to build some idea about what sort of like we're dealing with here in terms of raw numbers. And so please don't take this with a grain of salt and I please encourage you to try to justify and well around these numbers yourself. But number one said there's a 770 billion dollars lost to consumers by fake engagement. And so a fake a fake review on Amazon if you give a fake review to a shirt that's bunker fuel into the ocean just for a shirt to land up in a landfill. Like that's real

economic cost that is associated with fake engagement. So raining this in is just really important. So but if you don't want to deal with kind of like those like broad abstract numbers, we can see that like the bot defense market is about a half a million or half a billion starting in 2024 projected to increase by a couple order of magnitudes. Those are people just saying my service won't work unless I have something like Cloudflare. Um people always say that the market for defense is always a couple magnitudes lower than the market for offense. You can see that in Clear net botnet services which are 5.7 billion projected to grow running their ad campaigns. You know, you probably have seen massive armies of Twitter clones that have slightly

different names on them saying the same thing at the same time. Um that as as we are aware, you know, this is got to be the tail end of the number cuz we won't get any really good information out of those Southeast Asian kind of like swarm form phone swarms or anything like that. So, is it fair to say that by 2030 there could be $1 trillion lost in miss in capital misallocation because of this fake engagement problem. That's a lot of six-figure salaries for people.

It's a hell of a lot of six-figure salaries for people. And so, what what we should do, and this is why I'm going to bring the argument here. The economists call it a Pigouvian tax. If someone sets up a factory next to your house, it's really noisy, you want to hey maybe pay them off so the noise looks a bit like profit. And so, all we have have to do is say, "Hey, proof of work was originally created or originally envisioned for email anti-spam. Like, what if we just start saying 'Hey my attention is really important. They need my attention basically more than I need to put my attention online.'

Let's take the book page out of Judo, this Kuzushi or this gentle way. This idea that we can yield to opponent and use their own force against them. And so, maybe the Jedi master in a sense would say, 'They need to out-shout us because they need to corner the market on culture because it is the market of markets in capitalism. If you win the market on culture, you win capitalism. So, let them trip over all of our whispers.' And so, I like to bring a bit of an economic model. It's maybe a bit out of place, but it provides a good illustration. We have on the x-axis the

number of posts a user does. So, we can see we have a user posts how many numbers, and this is the cost associated with posts. We have on on the on the side we have posts or post bundles, and we'll see that when we compare a normal user to a flutter user. And we'll see that flutter users have a different utility than a normal user and this is what we'll exploit. So, the lower graph we see here is saying we have a social network and that's the sum total of all posters. So, we have maybe five posters on the network each of them contributed maybe 20 posts and so we have 100 posts total and it cost them $6 cumulatively to post.

The flutter who says actually my strategy is not relevant unless I capture signal dominance on the platform. It is simply not enough that I trickle in my data. I have to swash and quash all other knowledge information for my strategy to work. If one bad knowledge leaks, it's over. So, they need orders of magnitude scale up. So, we can see if our users post cumulatively 100 times, if they want to maintain a flood factor of say 10 to 1, they'll need to pay $60.

And so that's that's pretty interesting, you know, it's it's it's to go from 1 to 10, you only need two sets of hands. But to go from 10 to 100, you need 10 sets of hands. And so if we're getting a bunch of spam by just posting on chain and we're say oh well, let's just move to a more expensive chain. And then we can just really grab that order of magnitude there. So, while us collectively of users are paying 60, now they're paying 600. And so we can really see it just as it scales up that we just it's just exponentially growing that difference between the what the flutter has to do and what the users get to collectively do.

And to help build the case of why they need that order of magnitude scale up, I'd like to draw attention to the idea called containment control. Containment control was derived out of Eglin Air Force Base in response to the Arab Spring riots where they eventually had to cut the undersea cables if I'm not mistaken. But what they found is that information spreads like wildfire. And so if you have one bad person that's well-connected, it's game over. They spread it the signal will spread. And so, and this is why all of the major networks have moved to speech equals not reach.

It's cuz it's if you can cut the network at propagation time, you don't need to spam the network as brutally in order to hide information. Uh so, feel free to take a look at that paper. It has a lot of interesting mathematics if you're interested in math. And if you have more of a quantitative mind, uh the D-Wave uh structural imbalance of a social network uh illustrates really well mathematically about how they're thinking about cutting social graphs and sorting people into persuasion targets. And it's my hope is that after reading these papers, one is thoroughly convinced that it's like not they don't need just a flood factor of 10. If we meaningfully bring back speech equals reach, they'll need a flood factor of maybe 10,000 or 100,000. And that

translates directly uh into TX fees. And so, if we're going to play that numbers game again just on that on that model, and we're saying, "Well, if there's a million posts per day of a social network that we intend to capture by going with our strategy here." And we say, "Okay, well, if it costs 2 cents a post, you know, cuz we just add our 2 cents, you know, it's a nice little thing there." Um and we assume that maybe 900k of those were bot spam, just that 10:1 flood factor. And so, we have 2k being paid out by users a day with 18k being captured uh by transaction fees for the swarm.

Now, in economics, they do this thing called price demand elasticity. It asks this question about, you know, if this much more was demanded, how much the price would increase. Uh I got this off of F3 research, just going to assume this number. And so, I calculated and and please check me on this. It's was a confusing calculation, um that I could expect that these moving the image boards on chain would boost gas demand by 0.86%. Now that might not sound like much and if you assume zero correlation with F prices then that won't change anything with the actual F price.

But if you assume perfect correlation with F price to gas then this would be a 1.7 billion increase in Ethereum market cap. It sounds crazy but this is this idea that the million posts are being created every single day in perpetuity until till the end of the like future present value exchange rate or something like 20 years whatever. Um or a bit more concrete down to earth 7.2 million dollars per year in TX fees and that's being asymmetrically paid by the flooders instead of the actual users. So if a user decides to say hey I'm just going to stake my ether and maybe post once onto chain

it it it it introduces this weird dynamic where there technically theoretically exists an amount you can stake on chain and that you'll eventually go positive terribly said but like it's an invest post model where this idea is that if you get more people posting on chain more will be put into TX fees which would make the token price increase which would make the value of your stake go up. And so this idea that if you get more people just rallying this way is [snorts] TX fees for everybody. And so that I hope that draws the economic case for the world computer. I'd like to spend the rest of the time talking about actual um decisions that we can make in our dapps architectures

that enable this reality. Um and I'll do that through um a case study called Hash chain. Now the open access internet uh was chosen because it is uniquely situated. It is the first to fall in terms of the swarms of of anyone follows of that sort [snorts] of thing they were subjected to a hack two years ago. They were brought offline for a month. Uh people didn't know if it was going to come back. The user base never really came back. Um they had to deal with things like 15-minute captchas to talk. It's been absolutely degradation of the experience. Um it's a shadow of what it was. And then it absolute shadow. Um Um and it and indeed it's increasingly unmonetizable. I I think that bears just no reasonable explanation. So it lines

up having no um updates for 9 years on their servers. Um And yes, of course, like the existing cool kid doesn't align. Just having Cloudflare in there, if people remember back in the day, Cloudflare has had a really tepid relationship uh with with this uh anonymous community. Uh so that that's just not good right off the bat. Um and there was a hope originally that when we had AI that that suddenly they would come in and bring really excellent context. Like we have philosopher AIs in all these nests like sorting out everything like catching everybody. But what they found was put up by Crest Research is that it was a lot cheaper and a lot more effective to them just to fill these spaces with

drivel, content that looks genuine, yet is meaningless to hereby reduce their sense of community. Um I Images of from the Agora and the rating uh of the Library of Alexandria comes to mind just with lots of shouting of bots. Um so I think it's a really useful quote to remember here. Sometime it gets passed around as Descartes, sometime it gets passed around as Voltaire, but it's any community that gets its laughs and kicks up by pretending to be idiots would be eventually flooded with actual idiots who think they're in good company. Uh so if you're flooding these spaces with like uh this kind of like idiotic drivel, eventually it's going to attract all the idiots and it's going to destroy it. Um So

I like to take you guys through the story of a user going onto the service, using the service, and then leaving the service just so we can cover the stack on detail. Um I call this architecture a central serverless just as a just as a moniker. Um so you can take Twitter central serverless, you can take YouTube central serverless, uh take the image board central serverless. Um so it all starts with going to that DNS record. And I really do want to get rid of that. I'm much more happy with this universe where we can pass around a front end in a dist file or better yet just say, "Hey, don't trust my front end. Either use just Etherscan or ask even like your AI just to create a minimal interface."

Um some people have issues with this cuz you don't get versioning on it. Um I retort to that saying, "I don't want versioning or force updates. Everything should be backwards compatible." Um or they say, "I can't reverse proxy to my back end." That's the point. We don't want any back ends. Next I don't still don't know how to pronounce this guy's name. Please Sasbo said, "Thought third parties are security holes." We got to get rid of the black box central server. And if we decide to just put our bare-bones schema on Ethereum, then we can leverage that as the initial data layer.

And because invalid transactions are not propagated by the mempool, we get a horizontal scaling towards DDoS. And I I don't know if I can really understate how important that is for the current state of the net. Cloudflare currently controls 50 to 60% of traffic. All of it is SSL MITM. And we don't need it anymore. So we can remove the remove the need for Cloudflare or other third parties. And in addition, maybe as some sort of a side, we get archiving and anti-shadow banning guarantees. And so we'll leverage Ethereum to prove cryptographically that our message is guaranteed to be propagated the same way

at the same time across the planet at block speed. Uh if one becomes familiar with containment control, if one posts to Twitter right now, um you could it might get like two weeks from now. And um, you thinks you may think it goes towards your followers, it only gets sectioned out to a subsection of your followers. Uh, every piece of your digital likeness is portioned and sectioned out. Um, and just distributed towards their profit. Um, and so we can we can break that as well. Um, so a person now, maybe they got through the DNS record, now they're like on local host 3000 talking to their own local node. We're going to talk about toggleable identity.

Now, a lot of people in the Ethereum space especially like to talk about zero knowledge. Um, I want to actually and how we can maybe do a one or many relationship with identity. I think we can do a bit better. I think we can offer digital experiences that offer zero to many relationships in identity. Um, and so and we can do that by allowing users control how they fund their wallet initially. Um, so it's like they can come in with no identity, they can come in with some identity, they can come in with a lot of identity. Uh, so if you wanted to create like a Reddit service, you can demand their users uh, to have like a ENS record with a year's worth of posting history on chain.

Uh, if you want to do something a bit less, you can demand that they fund their wallet for a KYC exchange. Maybe that's about equivalent to what we see today. Uh, down below that non-KYC exchanges and then chip mixing from a fresh wallet. Uh, if you want more anonymity on that side. And so maybe now that we've logged in in essence with our toggleable identity, we got to talk about how we're going to fetch the data. And I am really keen on doing Uh, some people call it the dirty solution, but I think it was the original solution. Um, and that is client-side event log indexing.

Now, it is really unfortunate when the JSON RPC spec was made for Ethereum, I think it really came from the perspective that the node operator was the node user. Uh, that's no longer the case. I think most of us just use RPC providers. Um, and that distinction really came on the event logs. Uh, because it's really expensive to grab event logs. and that expense doesn't matter if it's your own node. Um, it's a lot better than it was today uh, or than it was 2 years ago, uh, but we're still we're going to fight to get this dream a reality. And it's a lot easier. For example, if you can fit a client-side index to your use case. So, so for here, for example, if you wanted to create a board that has a thread that has posts, you can simply just grab the uh, filter logs

for the thread and uh, for the thread and get all the posts, or or you can just grab the ID for the board and get all the uh, and get all the threads. Uh, so in this way, like you can just kind of go through the different web uh, you can go through the web experience. You don't have to grab all the indexes at once. As soon as they go on to one thread, you just pull the thread uh, from the event logs. Um, and this this makes a lot more sense, of course, uh, for more information technologies. You don't want to be doing such risky things with things like payments, right? In case the data is bad or becomes polluted.

Um, So, and like once we grab it from the event logs, let's use um, just in-browser a key-value store uh, to just to store it. And that provides a really nice uh, launching grounds for sharing this sort of data. And if anyone's curious about this kind of state with event log indexing, because I think it really is something super important in terms of Ethereum growing up, that client-side event log indexing becomes powerful, uh, to look at this uh, the Sebastian Burge talk at Pinos Areas called The Missing Punk, and it talks really well about what problems have to be solved for event log indexing.

Uh, the next part is because we're not getting any photos off the chain, it's just too big. We should talk about how dynamic media is going to happen. Uh, we can just go through IPFS. Uh, the part I want to talk about here is that the important part is about managing peer and platform liability. And because of legal content, spamming and DCMA is just a really important lawfare technique to take a site down. Like no countless of online services, they find out an admin has 12 hours to moderate, they just do 13 hours of content damage. And so if we can find a way to make our platform non-liable for the content, then we should go for it.

On the flip side as well, we've seen protocols such as Euronet back in the day go down this route, but then they shoved the liabilities onto the users. And so you're responsible for downloading all this random image content. And the best they could really come up with was to say just put it in an encrypted folder and no one can see it if you downloaded bad stuff. That is just a non-solution for a lot of people. So what we can do is just demand hot linking or self-pinning. And so a user can just basically host it for themselves or ask some other host service to host it. And then the and then the website would just grab it from there.

This makes liveliness an issue, but then liveliness can come in by people volunteering to pin other people's content to create more liveliness. And if you're looking for something that has 100% liveliness, then some sort of incentive structure could be built upon on top of it to get it to work. Um And so now that we have the dynamic images, we're going to talk about, you know, how we start filtering down those images. And it's it's funny cuz Blue Sky did a good job with their pluggable moderation services, but they didn't to best of my knowledge hard code theirs into their own. And there's as well as aspects of that stack that make it really impractical for a random person to run. Uh if you can set up an Ethereum node, you know, you're good to go on all of this, right?

And so what we're doing here is we're doing pluggable moderation services. We can basically create this idea that we're going to create a unified append only uh message logs, but we're going to filter on top of it. And we're going to let people compete for the right to choose. Just because the moral hazard of sis of ad of a cuz the moral hazard of a moderating in these spaces is just simply too high. And they can do this and we can make this efficient and actually usable by doing signed messages over WebRTC. And so it may be unattractive for a user to have any sort of bad content come up on it and then be taken down. Uh what a moderation service

can do if they're looking to be really greedy about it is just watch the mempool and then just filter on the mempool. And because it'll always be quicker just to sign and distribute the message and then go through the chain, you can have a really high guarantee that you won't even have if you bake your moderation service into your uh serve into your service, that they'll never even see the unsavory data. And I like to preface this as well is that by linking all your data to an Ethereum address, it becomes a lot more fair use ready. Now, I I this conversation seems to have gotten further away from the AI companies over the years.

Uh but ideally if there's any sort of I'd say justice to the data theft that happened, not that I'm really asking for it. I am a pirate by heart. Um that they could simply just retroactively pay people through the Ethereum address. And that's a lot easier than trying to find a Visa uh credit card uh to pay them back. So it's a lot easier just kind of pay it forward that way as well. Uh so in conclusion, um the follow of Web 2 monetization birth this botswarm gulag where we're really just handed this facade and this illusion of a service and we're just there to train their bots and be trained by their botnets.

They broke in order to really do this, they broke free speech is equal to reach. And and this is our economic opportunity we can bring back. And we can judo their internally online shouting match against itself for industry and public profit. And we can bring this to life by making our web services central serverless. And by by utilizing the Ethereum tech stack perhaps in the way that Gavin would initially envisioned it as kind of this world computer, we can make that a reality.

Thank you. Any questions if anyone has? Yeah, so yeah, so HashChains been I guess live online for about two two and a half years now.

It's it's been a quite quite a journey. In essence, I've decided to take no shortcuts, no KOLs, kind of just every post that you see that mentioned HashChain was written by a human. Just to kind of see what would happen about it and it's been an interesting journey. And it's been a fun journey. I think that So, and I'm really hoping too that more people would kind of see this as we get into the dead internet and I say let's make the decentralized Twitter, let's make the decentralized YouTube. I just started with the four the the old chance and everything it just made the most logical sense because they're the ones that are just in the fire and flames and yeah, without anything.

And always welcome for any contributions. It's totally use free to use however you want license. So, if you want the code, get the code, run with it. So, >> Okay, I I want to thank you again and really deep dive. I I I picked some interesting points there, but we'll talk to you after this and yeah, thank you and we'll close off with the this speaking session. Thank you. >> Thank you. >> Mhm.