Can Worldcoin Fix The Internet's 99% Bot Problem? | Co-Founder Alex Blania

🎬 DEBRIEF | Ryan & David Unpacking the Episode

Transcript:

Alex:
[0:00] I think the general headline is everything that we believe will work as a proof

Alex:
[0:03] of human, we will also build ourselves and integrate into WorldID. So we're not like Orb maxis or anything. It's like if it really is about getting to scale and about being inclusive and about working in a world of increasingly powerful AI, I think the Orb is actually the only way to get there.

Ryan:
[0:22] Alex Blania is the co-founder of WorldCoin. Alex, welcome back to Bankless.

Alex:
[0:26] Thanks for having me. Excited to talk. It's been two years.

Ryan:
[0:28] It has been two years. almost exactly two years since we released an episode with you and Sam Altman. And I want to start by maybe asking you a question about this because two years later, I feel like my social media experience hasn't gotten better. It's actually gotten worse. Like the whole problem that you were talking about, which is like, you won't be able to tell who's a human and who's a bot online. You were right about that.

Alex:
[0:52] We're here.

David:
[0:53] Yeah.

Ryan:
[0:53] And it's gotten like really bad. And I feel like AI is making all of this so much worse. AI is pretty much past the Turing test with respect to text. I can't tell who's a human and who's a bot. So let me ask you, two years later, there's a lot of traction for WorldCoin. We'll talk about some of the stats. But why hasn't WorldCoin fixed the bot problem online yet?

Alex:
[1:15] I think it's a couple of things. First of all, I agree. I think the kind of the AI general bot problem on social and internet in general, I think it's going to be a thing. But like, even more importantly, I think we're still pretty early on that trajectory. Because I think if you now take like a very competent open source agentic system or something like that, which I think we will see probably later in the year,

Alex:
[1:41] if not early next year, I think you will kind of another 10x whatever we experience right now.

Ryan:
[1:47] So it's going to get way worse.

Alex:
[1:49] Way, way worse. Meaning like... Yeah, I think we're kind of heading to a place where way north of 99% of everything happening on the internet will be AI driven. And that's like, for like most things, I think that's going to be amazing. You know, it's going to be agents that do like all kinds of things for us. But I think for things like our social networks, these systems will be smarter than us. They will understand us better. They will know everything. And actually there was like one very interesting paper that I think is like important to reference is like from the University of Zurich I think two, three months ago where the University of Zurich tried to change the opinion of subreddits with AI.

David:
[2:33] The R change my view subreddit. So there's a subreddit out there that's supposed to be like here's an opinion. I challenge you other readers to change my view. And what happened is that people leveraged AI to provide compelling, but also oftentimes false answers. Yes. And not only were they extremely effective, but one of the reasons why they were effective is because the bots had the time of day. They had the patience and the ability to just write out long responses and actually engage with the people where like typical change my view posts on that subreddit go mostly unanswered. And so it was incredibly effective. I remember reading the story. It was incredibly effective at changing people's opinions. Right.

Alex:
[3:14] And they even targeted on the people that made the comments. So like they kind of the AI was kind of tuned on the individual of like, what's their political view? How they did? What are the things they talked about before? So it's like, it's very targeted towards you and can change your opinion. And I think that's like, that is something that will be pretty scary to us, I think, in the coming years where, These systems will be smarter than us. They will know everything. They will be targeted to us and can change our opinion. And I think we will really need a proof of human for many of the things that we care about, like social networks.

Ryan:
[3:46] So is that what you guys mean when you say way worse, like it's going to get way worse? Because the way I experience it right now, it's more of an annoyance. So what I see is a lot of slop, I would say, on my replies on social media.

David:
[4:00] And you kind of know it's AI.

Ryan:
[4:02] You kind of do, or you're like, oh, that's like a low... Actually, sometimes I don't, David.

David:
[4:06] Sometimes the reply guys are that dumb.

Alex:
[4:09] Yeah.

Ryan:
[4:10] A lot of times, honestly.

David:
[4:12] The Turing test is being passed at the 80 to 90 IQ level.

Ryan:
[4:16] Right, but so it's slop, it's annoying. It's almost like getting spam in my email. I feel like there could be some auto-filtering out. That feels annoying to me right now. But when you guys describe 10x worse, I guess maybe I'm going to get the picture of somebody I've developed like a long-term relationship with. Someone that I know online and follow and maybe do business deals with. And it turns out like this person who I've been working with for the past year and interacting with and reading their blog post was actually a bot the entire time. That's a way I can see it becoming worse. In addition to me being targeted, like just being fooled at that level for a long period of time, that's very, I don't even know if I can picture a world like that.

Alex:
[5:02] Yeah, exactly. It will be accounts that just like build up interpretation, essentially. They just like post over time, interact with users over time and will really look like, it will not be slop anymore because it will be very somewhat sophisticated answers.

David:
[5:15] Sophisticated slop, yeah.

Alex:
[5:17] Sophisticated slop, yeah. So you will just not at all be able to distinguish anymore or what is real and what is not. I think that's kind of the default we're currently heading into.

David:
[5:25] I feel like this begs the question, is Web2 just cooked then? Like if Reddit, Twitter, all of these things are not equipped to manage bots.

David:
[5:36] This is, they're just not built under that construct. So like if these bots actually get effective at what they do and sufficiently prolific, where I think you said 99%, like it sounds like Web2 just, you can't operate under that environment.

Alex:
[5:52] We know from some of these platforms, they're really racing to try to find a solution. And some of them we were talking to. So like, it definitely is becoming top of mind. Of course, like the different platforms have different kinds of problems. For example, like Reddit is very, very far on the anonymity scale, where you basically by default interact with someone you don't know. And there's like, it's very comment heavy versus, for example, Snapchat is mostly your friends. And so like these, these two platforms have like very different shape problems of what kind of AI will do to the platform. So in any case, I think like that moment is here now and I think it's accelerating. I think to your question of why is world not solving that yet is it's a matter of of scale, obviously. I think like we're like it's a very, very hard problem to to solve. And I think us being able to, I don't know, serve 90 percent of the users of Reddit or even 20 percent of the users of Reddit is going to it's going to take some time. And we recently launched the U.S., which is going to be a key to that. And like we're still only, we have 200 devices across the U.S. So it's like, it's pretty hard to verify yourself right now in the U.S. And so we just need to really, really ramp up operation, get to scale and allow people across the U.S. And then eventually across the world to verify within like five minutes wherever you are.

Alex:
[7:14] That's like a big, big, a big challenge for us right now.

Ryan:
[7:17] Okay, so you think some of the reason world hasn't solved this yet is because you haven't scaled up yet and i guess maybe the my question is what what are the bottlenecks to scale is it literally the number of orb hardwares that like exist in the world and that are deployed is that the scale limiter are there some other scale limiters that you think of when when you say hey we like we still have to scale and we're bottlenecked

Alex:
[7:42] Yeah it's it's actually different things at different points in time like for example in the previous administration we did not launch the us just because given sam is a co-founder and like the stance of the sec back then we're just like very very cautious and so yeah if we would have launched us two years ago we would have been at scale here already but now that we're here it's just a matter of actually just like ramping up scaling up you need to currently we're actually manufacturing constrained so currently we just like produce much more orbs that will start hitting us in september end of september so currently we're hardware constrained at some point we will be deployment constrained so like you need an operations team to actually make deals with like i don't know chains across the us like maybe you want to have an orb in every walmart or in every clothing store or every convenience store and so like currently we do a lot of that bd to kind of get these devices, in there and roll them out that way. So it's different things at different point in time. And then we always go viral in different countries. And it's like, it's hard to predict. Like currently it's like, I think in Thailand, we are at a K factor of 1.7 or 1.8. So it's like, it's going pretty ballistic in Thailand. Early in the year it was Argentina. Last year it was Spain. So it's like, it's always... It's always different things. So it's a pretty complex network at this point.

David:
[9:10] The numbers that we can see out of WorldChain are pretty impressive. For those that aren't paying attention, 1.5 million daily transactions on WorldChain coming from 20,000 to 30,000 daily active users. Today, it's fun to be able to go look at these numbers. Today, there's been over 31,000, 31,043 users. In regular crypto world, non-WorldChain, non-WorldID crypto world, when we see a daily active user number, we all kind of just like reject it, right? Because we cut that number down by like 50 to 75%. So like, if you tell me some chain's got 30,000 users, I'm like, all right, you've got like 3,000 users at best. But on WorldCoin, on WorldChain, this is different, right? But is it one-to-one with a human? Like, really? Like, what does a WorldChain daily active user actually mean? Is it actually a one-to-one relationship?

Alex:
[10:05] So, well, technically, you could, of course, use the chain without your WorldID, because the chain is just an Ethereum O2. Chain is permissionless, okay, yeah. Yeah, the chain is, like, you could just, like, send transactions on there.

Alex:
[10:18] What practically is happening, and that's actually, that's a big focus for me this year, is, like, what we, of course, like, what we are building has to be a protocol, and there have to be many other clients that you can get your WorldID through, you can interact with world chain with and that's like practically just not the case, we're like right now pretty much all the usage happens through world app like because that's what we launched with and that just went viral and like that's the best product right now to kind of to verify with and use a world id with and so pretty much like everything that is happening on world chain right now is a result of world app and.

Alex:
[10:53] World app is

Alex:
[10:54] Just pretty large it's like it's now 30 million users in total 13 14 million of which are verified so like it's a pretty large app, And so like the numbers you just quoted, they are pretty much 100% rolled up. And so like these are real numbers. We know many of these users are verified and most of them come through a single client. And these numbers will continue to go up throughout the rest of the year. But the big focus is really to allow kind of many more clients and kind of really turn into kind of true permissionless protocol, which in our case, again, is like is hard.

Alex:
[11:28] But we will get there.

David:
[11:29] So how many people have you actually orbed? And then how many of those people get converted to world chain transactors?

Alex:
[11:37] So like 30 million have an account in the app. Then roughly half of that verify. So like we're now at 14 million that are verified with an org.

David:
[11:48] 14 million people have looked into the hole and gotten their eye scanned.

Alex:
[11:53] Yes, exactly. Exactly. And then of these, we have nine to 10 million monthly actives. So that kind of transact and role chain do things. So that's the rough ratios.

Ryan:
[12:07] So I am one of those users, I've not gotten verified, which David called getting orbed. And getting verified means I think you have to go find one of these hardware orb devices. You have to stick your eyeball inside of it. It takes a picture.

Alex:
[12:22] You don't have to stick your eyeball.

David:
[12:23] Just look into it,

Ryan:
[12:24] Yeah. Oh, really? Just look into it?

David:
[12:26] It's our arm.

Alex:
[12:28] It's like, actually, I think it's a cooler experience than Clear at the Opera, for example, because we just invested much more into imaging. But you just stand in front of it, and it's just like taking a picture. It's very fast.

David:
[12:39] Okay, so maybe... You don't have to do like this, right?

Ryan:
[12:43] Yeah, see, that's what I... When I look at the device, when I look at the actual orb itself, I get a picture of like going to my optometrist. You know those things where you ever get your cornea checked

Alex:
[12:51] And they shoot.

Ryan:
[12:52] A burst of air into your eyeball?

Alex:
[12:55] No burst of air anywhere. No, you just stand in front of it. It's like at the airport, like clear or something, or even just the border control. And you just stand there and 20 seconds later, you're verified.

Ryan:
[13:06] So I've got the app itself open. This is a world app. One of the, you know, I see all the world IDs out there. So you said 30 million. I could see all the kind of the accounts. I'm playing with this right now. I see the world coin price. And then there's this section called mini apps. This is kind of curious. This reminds me of like, you know, Apple, Apple store or something like this. And I've got some top apps here. There's an app called add money. There's an app called Uno drop wallet. And then I've got these, you know, gift card type apps. And then I've got this whole category, which is somewhat interesting called human only. So the idea, of course, with WorldCoin and the WorldApp is basically once I get verified, then you know in a civil resistant way, the app does, the network does, the protocol does, that I'm a human, I'm not a bot, right? That's the whole thing that you're doing here. That's the entire point. And then you can create classes of apps like human only apps. Maybe just talk about this whole mini-app idea. Is the goal for World to essentially create kind of an app store type of experience? And once you get people into the World app itself...

Ryan:
[14:13] Then you get them into the app ecosystem. Is that how you're going to market? What are all of these mini apps and what's the strategy?

Alex:
[14:20] Well, so kind of at a 10,000 feet level, the idea, I think, that I think back then we already discussed two years ago is that once you have that proof of human property and you're actually civil resistant, you can use the token to bootstrap that network to kind of critical scale. So let's say, I don't know, 100 million users is roughly what we think it will be. And so you're civil resistant, you use a kind of a token antrop as an incentive to get the network going, and then you build out utility for the network as you go along. And so that was like the very, very early strategy. And as the app got bigger and bigger, and we talked to a lot of developers, what they mostly asked for is like, how can they get this distribution? How can they get people to kind of ship their products to.

Alex:
[15:12] And that was the starting point for mini apps. It's like we have a lot of users in WorldApp. We will not be able to serve every use case. And how do we just make it easy for devs to ship something to all these users? And I think what it will end up being is that I think in a, if we talk again in a year or so, is that you will have mini apps for more so like a launch pad for kind of startups, devs, and kind of companies that are just getting started and them using the distribution of WorldApp to kind of get a product going. So for example, kind of something that is really picking up right now is micro-lending. So there's many apps called Credit.

Alex:
[15:53] I think there's still waitlist only, but there's a lot of landing apps that are picking up right now. And so they use the distribution. And then on the other side, you will have large companies that actually use WorldID, the proof of human. You'll probably have some of the social networks. You will have.

Alex:
[16:09] We had an integration with Match Group, Tinder. And so like, yeah, I think these will be the two sides. You will have like apps with a large distribution themselves that use the proof of human, use WorldID, use other features of the network. And then on the other side, you have developers that just need distribution and use mini apps for that.

Ryan:
[16:28] Okay, so let's talk about maybe the external. I want to come back to the internal with kind of the mini apps. That's why I'm calling your app ecosystem. I'll call that the internal. The external is where they're using, you know, World ID outside an external app. You mentioned Tinder, which is just like, are you telling me on like Tinder right now, there's like bots posing as like your next relationship and like that's a problem for Tinder? And Tinder can tap into World ID and basically prove some humans on their app?

Alex:
[16:55] Well, the short answer is yes. The longer answer is one thing that we believed in pretty early on is that it's not just that you will need to have a kind of a proof of human, but you will need to be able to prove things about yourself. So for example, when you verify for world ID, you actually get a signed face image that the orb captures and you have that on your phone in a self-custodial biometric package. And you can then use that to prove things to apps. So for example, you can prove to Tinder that not only that you're a verified human, the individual that is actually shown on that profile picture is you. Or you can log in into Zoom, and I can prove to you on Zoom that what you see right now is actually me because there's a real-time comparison to my world ID package. And so these are all the things that are rolling out now. So it's like you will be able to do video calls. We call it product DeepFace.

Alex:
[17:53] So you can prove

Alex:
[17:54] Yourself on video calls. You can prove your profile pictures on dating apps, for example, on social networks. So it's not just I'm a human, but also I am who I claim to be on Tinder. And then the next step will be you will be able to actually like prove specific actions. So like I will be able to send you a message and be like, hey, on Tinder, you just asked me to meet in person. Is that really you?

Alex:
[18:20] And you have

Alex:
[18:21] To prove like it prompts essentially like a face ID check on your phone to agree. Yes, that was actually me. and let's meet in person.

David:
[18:32] Is this a problem on Tinder?

Alex:
[18:34] It is a big problem on Tinder, yeah. There's like a big catfishing situation going on where like there's AI created profiles all over the place and they kind of catfish of like whatever.

David:
[18:50] Like what are they trying to get from their targets?

Alex:
[18:53] Well, there's a whole range of things. It's like it starts with simple things like phone numbers or gift cards to actually convince someone to go to a date and then on the robberies to kind of even worse things. As if the dating world.

Ryan:
[19:11] Wasn't complicated enough. You have to worry about this now.

Alex:
[19:15] It is a thing. It is a thing. And I think like the, but that is like in general, I think the general statement that human interaction on the internet will require a new version of trust. Like Zoom calls, there's all these news now of CFOs of companies getting scammed because you had a Zoom call of the CEO calling the CFO and tell them to wire something, but it was actually a deep fake.

Alex:
[19:43] So dating, gaming is another big one. So we integrated with Razer, which you might know from the gaming hardware, to kind of bring Roll Daddy to gaming, where bots are now obviously better than professional gamers and like most of these games and if you, play for hours a day and you train to be good at a certain game and then you actually you play against an ai that's that's really annoying and there's a lot of money involved in the business too so it's like it's gaming dating social all of those different industries that we currently get into.

Alex:
[20:16] Amazing. Yeah.

David:
[20:17] So there's potential MMORPGs out there that can gate their app with a world ID so that we have a human verified only MMORPG that

Alex:
[20:29] We can play. Yes.

Ryan:
[20:30] Correct. I think for me, the two things like maybe stopping me from getting like fully orbed and going all the way here are like maybe the incentive to get orbed. But I would call this like number one, there's some monetary incentive. You get world coins in order to do this. I think when I checked in my app, it's like $50 worth of world coins, right? I'm like, well, I mean, if that was $5,000, all right.

David:
[20:56] If that was an airdrop size,

Ryan:
[20:58] Yeah. But for me, $50 is not quite enough. But what can compensate for that is if I'm getting utility on the other side, right? And I don't yet know in the mini app ecosystem you have internally to the world app, whether there's anything there that would provide me enough utility to go do that. The other piece that's kind of stopping me is like privacy. And that's a separate conversation. Let's stay on the utility piece. One thing to me that would probably push me over the finish line in terms of utility is if the WorldID was like integrated in some of the commonly used social media apps that I use on a day-to-day. So I'm not a Tinder user. Bankless listeners will know. It's a long time since I've been in any of the dating games. So it's not a problem for me. But on Instagram or Twitter, I mean, like X for me is just like if we had some sort of proof of humanity on X, then that would be a game changer. So I guess my question is when you think about scale, you're thinking about the internal mini apps and then you're also thinking about these external apps. What's the plan to get WorldID integrated into all pockets of the Internet that a user like myself might locate to add utility that way?

Alex:
[22:10] Well, we're working really hard on that. Just a short answer. And then the longer answer is like, I mean, a year ago, so two years ago, people just made fun of us. Three years ago, everyone made fun of us. Two years, like a little bit less people made fun of us. A year ago, people appreciated it might be a thing, but it's not yet. And so like people didn't want to meet with us. And now suddenly like all the big companies want to meet us and figure something out. And it's still like on the cusp of things where like you have some of these social networks that really already have the problem, like X and Reddit, I think, are two examples of that. Snapchat, not really yet. And so it's still ramping up. But kind of my short answer is, yeah, I would bet that we will be integrated with many of the services you care about in the next 12 to 24 months. I think then it will still not be the case that a large fraction of the users of these platforms use WorldID. So it's more going to be then a status thing of like you will have a human badge and maybe like she will get.

Alex:
[23:13] You will ranked higher and kind of in the ranking your comments will show up earlier something like that and it's then still going to take us time to kind of really get to i don't know 10 20 30 40 percent of these platforms but i think like the early the actual integrations i think are about to happen pretty soon for many of these platforms so in my mind all of this is an equation you have like utility which is like the integration first they need.

Alex:
[23:37] To exist and second is what does the product of these integrations actually look like so for example i think for people like us that are maybe in tech and use these products a lot i think status will be the will be actually the kind of the biggest pull initially like maybe you're able to do things that other people are not you have like you can prove things about yourself that other people cannot so i think that's going to be a big pull and so these product integrations how they look in detail really matters and then the third bucket is we need to bring down the friction for you to verify by a lot so for example like if you could imagine and in Argentina and in Buenos Aires you can already do that like if you can imagine let's say a world that he would be integrated with x and then you get a human badge and whatever like you get a kind of verified check mark and maybe your comments are ranked higher you post the rank higher something like that some combination of different things but then also you can verify by with uber essentially like uber eats kind of situation where the device actually comes to your home you get verified you get your world id and you don't even have to go anywhere so like if the friction and so like that actually we do that with rapi and latem so you can just literally order an orb via the rapi app and get verified and so like, there's many many different things coming together the friction needs to go down to zero So integrations need to happen. The product integration needs to be great.

Ryan:
[25:03] Let me ask a question about these platforms, the Web2 platforms that you were talking about that most users would love some sort of proof of humanity inside of today. What's their alternative to what you're doing with World, right? Is it basically like driver's license and passport, basically citizen ID, a lot of the AML KYC type stuff? Because that's from a practical user perspective on the internet today. That's the only other form of ID that seems to do something. And of course, that's like so easy to get. I mean, all of that information is just like everywhere. But I guess maybe you answer the question, what's the alternative to what you're doing with the world for them?

Alex:
[25:47] It would be KYC,

Alex:
[25:49] But it would actually, I think the most promising approach would be you use the NFC-enabled passports. So like some of the passports have NFC chips in there where you get the actual signature and then you could deduplicate on that signature, for example. And we actually, we do that in Roll.id. So you can get a Roll.id already today with an NFC-enabled passport just so we make it easier for these platforms to like scale out the users and it's not just Orb gated. Meaning like, I think the general headline is everything that we believe will work as a proof of human, we will also build ourselves and integrate into World ID. So we're not like Orb maxis or anything. It's like, it's just that if it really is about getting to scale and about being inclusive and about working, in a world of increasingly powerful AI, I think the Orb is actually the only way to get there after thinking about this problem for like a very, very long time. So I think there will not really be a way around this or something that looks a lot like that. It's like there's also now competitors that are emerging. A lot of companies raising money to kind of build a world competitor to build different kinds of hardware that are palm-based, face-based. Like I think you will see different, different offerings in that direction.

Alex:
[27:08] But the only alternative would be, would be government IDs that are, I think the privacy implications are.

Alex:
[27:15] Conceptually really, really bad,

Alex:
[27:18] Especially if you talk about social networks because I think like the freedom of speech, the anonymity protecting all of those rights I think is so critical. So I think it would be a horrible outcome for that. And then, yeah, I could go much more technical, but I think that's the high level.

David:
[27:36] I want to turn that conversation to the incentives because like Ryan said, you get about $50 at the current prices of WorldCoin. Do you call it WorldCoin? Well, we need

Alex:
[27:46] A bull market.

David:
[27:46] That's going to be one. Yeah. And so I would imagine one of the big reasons why people are getting onboarded into World ID is because of the incentives. I don't know if you have any intuition or notion about like, is that 50% of the people getting orbed? Is it less than that? More than that? I don't know if you have any sort of idea about that.

Alex:
[28:09] It is more than 50%, I would say.

David:
[28:11] That are coming to the incentives?

Alex:
[28:13] Yes. I think that's... And that was.

David:
[28:15] Always part of the plan. The plan was to like,

Alex:
[28:17] We're going to have to go over. Yes, I think that is the plan. I think it's going to remain like that for probably another year, is my guess. Where like now you actually already start, like some mini apps pull users in, some integrations pull users in, like you already start seeing that transition happening. But I think until you see that being the case for like, I don't know, everyone that signs up for World because like, let's say World is integrated into.

Alex:
[28:41] X and that's now

Alex:
[28:42] Kind of a core user flow in X, I think that's going to take another year at least.

David:
[28:49] And then in addition to that, there's also incentives to use some of the applications. And I think with Morpho, Morpho is deployed on the World Chain. And so you can go and use Morpho. But typically with DeFi incentives, they're TVL based, right? So if you deposit $10,000 into Morpho, you'll get some amount of incentives. And then if you deposit $100,000, well, you'll get 10 times the incentives. But that's also not how it works on world chain because you guys have world ID and so you guys are able to deploy world incentives on Morpho on an individual account basis which is pretty cool. Can you talk about just like because you have the world ID talk about how you can leverage your guys' treasury, the world incentives differently than anyone else can because you have this feature of proof of personhood. How are you guys getting creative with incentives?

Alex:
[29:43] Yeah, so first of all the Morpho mini app is actually a good idea is a good example where like I think on base there's around 3,000 depositors of Morpho and on world on world chain it's like 30 ,000 30,500 something like that so it's and essentially what the incentive is that Morpho's giving is and I think the foundation it's a combination of both is the first $20,50 that you deposit, kind of you get additional yield and we can do that because we have world ID so it's not it's not just whale based in some sense. And so of course it's a different user group. It's like it's everyday people that deposit their money into our more for mini app versus like some mega whales. But I think that's that's why we're building this. I think that's cool. And other than that like other than the actual core kind of token incentive like the core idea of like eventually every human will have some amount of world coin. We actually don't do that much. We give rewards to mini app devs that are kind of somewhat early.

Alex:
[30:43] Like we have a mini app

Alex:
[30:44] Dev fund. which we support them with. But that's pretty much it.

David:
[30:50] What is the size of the actual like war chest that you guys have? How much do you have of WorldCoin to actually hand out for incentives?

Alex:
[30:58] The overall tokenomics was right from the beginning. That's like pretty much 75% will go to users. Okay.

David:
[31:05] And the actual core airdrop. It has been a war chest.

Alex:
[31:07] Yeah. That is what we're going to, what are you going to use it for? Like it's, again, most, like pretty much all of that is going to be the kind of the core WorldCoin airdrop. So the simplicity that eventually every human will have some amount of these tokens so we actually don't freely just play with these tokens but rather, they're dedicated to when you sign up you get some amount of world coin and we will follow that kind of curve so and I think that a treasury of of the foundation at this point yeah a couple hundred million WLD okay yeah.

David:
[31:38] Yeah well at a 10 billion dollar fully delegated valuation the war chest is like pretty damn good one of the better ones we've ever seen which like the reason why I'm getting here is because this is also a bug bounty of the Sybil strength of World ID. Right. And so if there is this much money to be given out, certainly people are trying to figure out how they can get their hands on world coins that, you know, just not just theirs, but everyone else's that they can get their hands on. So how is that going? How has the strength of World ID's Sybil resistance held up? Are people poking through? How are you managing that? Tell us about this.

Alex:
[32:17] So actually like the last time we spoke two years ago, this was one of my biggest concerns. I was like, okay, if this, if the launch goes like amazingly well, this is going to be, as you just said, like a multi-billion dollar bug bounty on a new hardware platform. We actually did not see, we saw a lot of attacks, but the system held up against everything so far.

Alex:
[32:38] So that's, that was like very,

Alex:
[32:40] Very, very good.

David:
[32:41] When you say the system, What does that mean? Does that mean the orb? Does that mean the chain, the WorldID app? I mean, so like...

Alex:
[32:48] I think the biggest one was the orb, definitely, because that is where you actually could fool the entire thing.

David:
[32:55] Right, yeah, if the orb gets cracked,

Alex:
[32:57] Then everything falls apart, right? Still, the orbs would be rate limited. So even if.

Alex:
[33:03] You get one

Alex:
[33:03] Orb and you attack that orb and you succeed with that, it's not that you can just mint a billion WLD or something. And there is a spoof detection system going on at any point in time where if.

Alex:
[33:18] An orb behaves

Alex:
[33:19] Not the way it normally would, like just much more synapse or it doesn't move or things like that.

Alex:
[33:25] That will be detected.

Alex:
[33:27] And the cool thing is actually all of this will be in the public pretty soon. So you will have public watchdogs of people that can monitor the entire orb network and can spot if something is not working the way it's supposed to. But yeah, we had a lot of people try to attack the orb. We did not see any kind of meaningful breaches.

David:
[33:46] Okay. Orb is still holding strong.

Alex:
[33:48] Still holding strong. Yeah.

David:
[33:50] Okay. Where have you seen things break down and how bad have those breakdowns been?

Alex:
[33:55] The biggest one was like, honestly, WorldApp. I think WorldApp then before, at some point we were on OP mainnet before we launched WorldChain. And back then the combination of like a lot of user traffic and kind of a pretty high market sentiment. and they're all like led to kind of chain congestions back then and so like transactions would not go through which.

Alex:
[34:19] Was a pretty

Alex:
[34:21] Annoying experience on world app that was like a big part of why we actually launched our own chain just because i think we have been like 80 percent of the block space of op mainnet back then which.

Alex:
[34:30] Was bad for us and

Alex:
[34:32] Bad for everyone else in op mainnet so we kind of were forced to launch our own chain back then that was probably number one problem and two is like world app in general. I think we, now it's pretty stable, but in the early days of the launch, we had outages and like, again, just bad UX around that. These would be like the main two, like the chain itself and the app.

David:
[34:55] So if you had to give yourself a grade, if you had to go back two years and tell yourself, previous self, hey, here's how we did. We got A minus, we got a B plus, we got an A plus. What would you give yourself versus your expectations two years ago?

Alex:
[35:07] I think, I think, I think it depends on the, do you want me now to do a grading exercise.

Alex:
[35:12] On the company?

Alex:
[35:13] I think, let me think. So I would say on, I think on thesis.

Alex:
[35:20] I would give us an A plus.

Alex:
[35:21] I think like overall strategy and thesis, I think was an A plus for like, I think.

Alex:
[35:26] Again, back then, everyone was giving us shit.

Alex:
[35:28] Everyone was making fun of us. And I think, like, the core idea was actually pretty good and made sense. And I think the timing was pretty good. So, like, I think, like, thesis strategy pretty good. I think on hardware engineering, I would give us an A. I think, like, we could have been faster, but, like, everything is working. So, like, the whole, like, biometric pipeline, I would give us an A. So all of that system is working pretty well. On operation, I would give us a B to C, actually. I think we could have been much faster in the overall rollout. So it was a lot to learn there and like dealing with privacy regulators, dealing with governments. Like it was like a lot of stuff.

Alex:
[36:07] Back then to go through.

Alex:
[36:08] Overall engineering, like an A to B.

Ryan:
[36:10] Let's talk about maybe the other thing that's holding me back from getting my eye photographed by the orb.

Alex:
[36:17] Yeah, orbed.

Alex:
[36:18] Orbed.

Ryan:
[36:19] Orbed, getting orbed. The thing that's holding me back properly is privacy. And it's like truth be told i haven't done an exhaustive deep dive on world privacy right but this is important biometric data i mean there's all sorts of ways this has gone wrong in the past when you know data has been shared i mean like people have recently recent memory is this the past year 23andme where's that data who owns that i mean like who knows right You can even look at something like ChatGPT and this is not the self-sovereign data privacy that I want as a crypto user, basically, right? So internal employees under certain circumstances with controls can access your chat logs in AI. It's not just ChatGPT. I assume all of the AI chatbots are like this. Even when you turn it into like disappearing chat mode or whatever with ChatGPT,

Alex:
[37:15] That can all

Ryan:
[37:16] Still be accessed. Tell me about privacy here. And how can I, I don't want the trust, this is crypto, right? So we can't do the trust me, bro. Hey, we'll keep your data private. And like, if you get subpoenaed by a government, for instance, it doesn't have to be the US government, some governments more in the world. Like, does that government just get my data? And what does that mean? How do you answer my concerns about privacy?

Alex:
[37:42] Yeah.

Alex:
[37:43] So actually, I think like, this is the audience to, where all of this should resonate with because I think in all the other cases you just mentioned what of course has been missing is cryptography and that's why this space exists so I think like you can, many of these things are counterintuitive like zero knowledge proofs in general I think are still counterintuitive to most of the population but there's more things here we have like multi-party computation we have CKPs and I will.

Alex:
[38:08] Walk in kind of

Alex:
[38:09] In that direction in a second but like I we The kind of a general idea for this project right from the beginning, it obviously did not resonate like that in the beginning, but I think that will change, is that I think proof of human will become this very, very fundamental thing on the internet that people need to use for pretty much everything we do. And I want this to be the signal equivalent of like this is like the hardcore privacy alternative that is independent of governments, that is fully provable and decentralized and like that's actually what we're aiming at so we have a big big chunk of the team it's now a 500 people company so like a pretty it's a pretty big company it's a pretty big team, and a large chunk of the engineering is sitting on privacy yeah, And kind of the core idea is without not going too much into detail.

Alex:
[39:00] But is making like,

Alex:
[39:03] So you sign up with an orb. First thing that happens is it checks that you're an actual person. So there's not a display. So it takes a picture in multiple wavelengths, actually, infrared, near infrared, RGB. All of that happens on a device. And then you have neural networks on the device that check against all kinds of attacks. Then takes a picture of the iris. That is because that's actually one of the very few modalities that will get us to billions of users because face ID, for example, on your iPhone, that is a one-to-one comparison. So it just checks that you're again Ryan logging into your phone. But to solve the proof of human problem, we will need to compare you against, let's say everyone else that is using X and make sure that you as Ryan, you only get one X account versus like a thousand. So it goes from one to one to one to N in our case. And so face doesn't work, fingerprint doesn't work, like some palm veins could work and iris certainly works and so that is why why that then it calculates an iris code out of that picture on the device.

Alex:
[40:07] And it actually splits

Alex:
[40:09] That iris code into many different pieces and sends it to a multi-party computation so like right now there's the university of well it's berkeley friedrich like a german university which i don't matter got at fau, NetherMind, and a couple others that all have to come together to do one comparison. That means like any of the party doesn't actually have the data, but they need to come together for one multi-party computation. If that uniqueness is then given, so we know that you, Ryan, actually don't already have a raw data account and that MPC computation comes back positive, then public key gets inserted into a Merkle tree.

Alex:
[40:48] Which is not easier, it may not. And then how I actually use my world ID is if I go to a platform, I go to X, I prove to X that I'm actually, the private key that I have on my phone self-custodially is included in a Merkle tree. A public key generated by the private keys is included in a Merkle tree. And so, meaning you have the multi-party computation, there's no central storage of any information, but then the actual usage of world ID is decoupled even from that with zero-notch proofs. And so you remain fully anonymous using your world ID, and there's no database or anything that can get compromised. That was, again, there's like a lot of things that I think still can improve. Like, we want to have the MPC computation not be four parties, but like 50. And like many other things that I think can be tweaked and kind of improved. But in general, I think this is the most private, by far the most private solution we have for.

Alex:
[41:50] That kind of problem. Like for example, using,

Alex:
[41:53] Yeah, even using like NFC IDs from a government with zero knowledge proofs would not get you nearly as far in privacy scale.

Ryan:
[42:02] So Alex, if a three-letter agency shows up and they're like, hey, this Ryan Adams guy, he's kind of a shady dude. and we need you to give us all of the information, everything you have on Ryan, Sean Adams, what can you give them and what can't you give them?

Alex:
[42:17] Well, we can give them your transaction history, your financial transaction history, because it's an Ethereum wallet. So transactions are in the public. We can give them some app usage information. Files of like when have you been online what have you done in the app which screen did you use so there's there's things like that we cannot give them your biometrics we cannot give them your kind of where did you use your world id none of those things because all of it is self-custodially with you on your phone so so yeah i think it's like we we again app users just like the tlr of what we as tools for many could give someone.

David:
[42:59] So if ryan logged into some application that integrated world ID. So like he logged into how to build a nuclear bomb.com with world ID. You would not be able to report that to the A3 letter agency. Neat. No.

Ryan:
[43:14] Why'd you use that example,

David:
[43:16] David? I don't know, dude.

Alex:
[43:19] Because it's incredibly

Ryan:
[43:20] Unlikely that I would ever do that, right?

David:
[43:22] I don't know. Let's go to that website right now, actually.

Ryan:
[43:24] Well, just by the domain. Just on the privacy issue, what kind of privacy regulations have you run up against maybe in other jurisdictions. I'm not sure if there's any in the US, but I mean, EU has all sorts of things, you know, like privacy. What's been kind of, I guess, interesting or at odds with how you've designed things or like where things fit together maybe?

Ryan:
[43:47] What has your experience been with regulation, privacy regulation that governments are putting out?

Alex:
[43:51] There's like a whole scale of things that I think we encountered. I think like the first thing is like we launched back then and you might remember that, but it's like a very, very loud launch where I think pretty much in every country that we already were operating at or we operated before, it was like a national TV type situation.

Alex:
[44:11] And kind of in

Alex:
[44:13] All the big newspapers.

Ryan:
[44:14] Okay. Was it basically a Sam Altman's orb is coming for your eyeballs type of thing? Yeah. Okay.

Alex:
[44:20] Actually, it was like a headline to write.

Ryan:
[44:22] Yeah, it's a pretty easy headline. There's some lazy reporting there.

Alex:
[44:25] For better or worse, actually, like back then it was already much more positive. Like, meaning like...

Alex:
[44:29] It was like,

Alex:
[44:30] It was still Sam Altman is like doing X, but it was like, oh, Sam Altman's like crypto project is launching in Germany or something. It was like originally for Germany. Like on every major newspaper, it was like headline Sam Altman is doing this new thing and it's an economy. So it was actually, it was the reason that like the sentiment back then was pretty positive around the launch. But then the second thing that happened is like back then we didn't have a GR team. It was like me and my GC. Back then the company was like much, much smaller. And we had like within the span of two weeks after launch we had pretty much every regulator you could possibly think of reach out to us from like financial regulators crypto regulators privacy regulators like, like everything Is this just how it happens?

Ryan:
[45:15] Like you get a big media splash and then governments are gonna all the government agencies

Alex:
[45:19] Are gonna ask questions That is totally how it happens and for what it is like if I were a privacy regulator and I wake up to a newspaper saying there's like Sam Altman with orbs, in like all of our big cities and somehow it's getting eyeballs and cryptos involved. Like, of course, I would take a look at that.

Alex:
[45:38] Like, it's like,

Alex:
[45:39] That's the job description of a privacy regulator. And so, yeah, we basically got outreaches from like every privacy regulator in the world that we, even in countries we didn't touch. I was like, just like, hey, just FYI, if you want to launch here, we want to talk. Like, that kind of hundreds of emails. And then, of course, we also fell like squarely into all the crypto regulation debate back then where like Europe was forming their crypto regulation in Southeast Asia. You had many countries that didn't know yet what, what crypto regulations. And like me and my GC facing, facing all of the drama. And then the, the second, the second order was like, so you now actually meet with a privacy regulator and you have to explain them, multi-party computation, zero knowledge proves.

Alex:
[46:26] All of that

Alex:
[46:27] Stuff proof of human we will have AI on the internet and it's going to change everything it was like the, conversation was so disconnected from the normal conversations that it used to have that it really took us like months to just get through the basics and so the good thing is like every, like some of the conversations are still not concluded meaning we're still having conversations or still kind of going through because some of them really went pretty deep like they went for source cord or like hired consultancies to kind of go into things, but either they conclude positive or they will be concluded positively pretty soon. So overall, it was a pretty positive story, but of course.

Alex:
[47:07] A lot of work

Alex:
[47:08] To get there. And then actually like my favorite stories are the ones where like, backdoor government kind of conversations that are like, well, you would allow democratic kind of votes or something in our country. We don't like that. And so we would never want you to launch here.

Alex:
[47:24] It's like, at all kinds of crazy conversations over the last two years, you have no idea.

Alex:
[47:27] And like every scale. So, yeah, I have a lot of

David:
[47:31] Stories to tell. What's something that you hope gets built on the World Chain, World ID platform by like the end of this decade? And maybe this can be like a fun toy that you just want to use or maybe this is something that, I don't know, since you just brought it up, something that like disrupts authoritarian regimes or, I don't know, something exciting. What's something that you daydream about?

Alex:
[47:54] That it becomes like some version of internet citizenship where you plug in both your, you have some version of internet native identity and internet native finances, that kind of almost like give you an internet native citizenship And there's like billions of other people that have that. And it doesn't matter where you come from, you have like access to the same kind of tools and systems. I think that's the future where we're heading to. And I think that's like the cypherpunk dream to some degree.

Ryan:
[48:26] That's certainly been a bankless dream for a while, having internet citizenship. And you guys have come a long way in the past two years. And the way you're building it, I think, has been impressive, at least to David and I, as we've been watching this with kind of a layer two, doing everything with ZK, building it with privacy in mind, actually scaling it, 30 million users is pretty impressive, 15 million of which have gotten the orb treatment.

David:
[48:50] It's worth pausing and reflecting, actually. The topic of identity has been with us since getting into crypto. I think it's one of the, when you're going down the crypto rabbit hole and you get to the identity part of the crypto rabbit hole, you realize like, oh, an Ethereum address is an identity. And then it kind of goes from there. This was on Bitcoin Talk forums. Bitcoin addresses mean identities. and then it opens you up. It's like, what is an identity? And then we have to deconstruct an identity. And there were so many different ways to solve identities. There was Martin Koppelman's Circles project. There were so many different identity projects trying to solve identity.

Ryan:
[49:25] Nothing that scaled though.

David:
[49:27] Nothing that scaled, nothing that worked. This one has gotten so

Alex:
[49:31] Far down the.

David:
[49:32] Line of actually solving identity. And actually it doesn't feel like there's any unknowns. It's just an execution at this point on WorldCoin's part it's no longer just like a science experiment it's actually like in production and so it's actually I think it's kind of worth like pausing to reflect on like oh like I didn't expect it to come from hardware turns out the key unlock was hardware but there were so many other like attempts at building identity that never really went anywhere and this is actually the one that is so far showing plenty of fruit and promise of actually getting us to the internet citizenship which we all kind of daydreamed about so first

Alex:
[50:04] Of all thank you I think the we're always still very very early and it could go like, could fail for like a thousand reasons. But I think it actually has a meaningful shot at getting to kind of internet scale.

Alex:
[50:17] And I think actually if it does that, it's going to be much more than identity. I think that's the cool part. I think like the early vision of you create a real human network on the internet and you have financial primitives built in will become one of the most valuable networks on the planet, I think is fundamentally the right idea. And I hope we get there. And actually, to your point, one of the funniest things is actually how Finney had proposed proof of human in the kind of very early Bitcoin forum, that like eventually you would get to a biometric device and you would get a proof of human and you would be able to use it across the internet. So it's like a, it's a pretty, it's a pretty OG idea.

David:
[50:59] How Finney concluded biometric hardware?

Alex:
[51:02] Yes. Wow.

Ryan:
[51:03] One of the contenders for maybe Satoshi, who is Satoshi.

David:
[51:07] Yes, this is, when I hear Haltending, I'm like, okay, so you mean Satoshi.

Ryan:
[51:10] Oh, gee. Alex, this has been great. Maybe as we close this out, I want you to reflect, if you will, on kind of some convergence here that seems to be going on. So we have another podcast that David and I sometimes occasionally guest host called Limitless, and that's exploring the frontier of AI, the way Bankless explores, you know, crypto. We're seeing a lot of interesting developments, right? If we're looking at kind of the next generation of social media platforms, it feels like there's like four things that are kind of emerging as important there. The one is it's got to have money. And that's like a crypto wallet, probably. And you guys have a crypto wallet inside the app. The other is it has to have media, right? So you have to have the breaking thing. You have to have a bunch of humans congregating over some sort of information, whether that's prediction markets, whether that's traditional media, tweets, that kind of thing. The other thing is it has to have intelligence. So that's some kind of a model, right? So Grok is kind of X's attempt at this, XAI, and ChatGPT, of course, is the open AI attempt at this. And one wonders whether open AI is actually going to launch its own social media platform. And then the last ingredient, the fourth, is basically some sort of identity, some sort of proof of humanity. And it seems like all of these ingredients need to be active together in order to launch our next generation of internet experience, social media experience.

Ryan:
[52:28] And so I don't know how people are assembling the building blocks here, right? A bunch of companies building models. Obviously, we have the crypto thing in the background. We have World, which is working on kind of the humanity piece. I don't know if there's maybe some smaller competitors on this.

Ryan:
[52:42] All sorts of like, there's a milieu of part, like it's kind of a stew of all of these bits and pieces. What do you think happens next? Do you think all of this converges together and creates a better internet social media experience for us all? Or do you think there's like some balkanization where there's maybe the Elon Musk world, there's the Sam Altman world, there's the Google world, and these are all kind of separate using different protocols? What do you think is going to happen?

Alex:
[53:08] So first of all,

Alex:
[53:09] I do agree with, I think these are the fundamental building blocks for the next generation of such a platform. And I think probably right now is the opening to take a new shot at something like that. And so like, I think the timing is very interesting. I think the observations you made are pretty correct. I think the money part is probably initially secondary, but I think that the idea of like you have an AI native social network that somehow also is able to have kind of a positive coexistence of AIs and humans, meaning you will need some version of identity or proof of human. I think that will happen. I think the other big thing that will happen is there's going to be some form of flywheel where.

Alex:
[54:00] The kind of

Alex:
[54:00] The big model certainly the big labs will have it but probably other servers will have it too where like I think you will get more and more personal context about everyone by usage of these models, which then I think should allow you to build a much much better feat, a much more personalized feat that actually just shows you the things that you're interested in and I actually think like the, It's like, it's not, it's hard to say that X makes you a better person.

Alex:
[54:26] Is it?

Alex:
[54:27] Like, it's like, it's really hard to claim that like X is like truly just improving my life and it's making me a better person.

Ryan:
[54:32] I took a break for three months and I felt great.

Alex:
[54:35] Right. And so like, I think we all feel like that. It's like, it's like, it's like a mix of like the craziest shit on the internet all thrown at you all day long.

Alex:
[54:43] And like, and like some, some percentage of interesting things in between.

Alex:
[54:47] So like, it's not, it's like, it's really hard for me to say this. Like it makes my life better and I really really need that I I use it a lot because I need to use and then I kind of I take down all the negatives that come with that but I think my hope for the next version of such a such a product would be that because it understands me and it knows what I care about it actually shows me things that either just make me better or just improve my life, improve my experience versus like a hellhole of the craziest shit happening on the internet. Like that's actually nothing I need to know. And so like, I think the other big flywheel from the things you just mentioned, I think is personalization to, towards all of us.

Alex:
[55:34] And it comes with a lot of danger actually, because the, the potential downside of that level of personalization is that we actually don't have a public town square anymore, but rather we just like build dopamine maximizing systems that like they will make me very happy and feel great but actually i i miss out on what is happening in the world so there's like a balance you need to make but tlr is i think it will be vulcanization i think there's going to be just that's the nature of that i think these networks will get more and more personalized i think these networks will include financial permits that definitely will include some form of proof of human or identity otherwise they will break and if there is a moment to start a new major network it probably is now.

Ryan:
[56:18] Maybe I'm reading some things into what you just said there, Alex, but maybe it sounds like there's some announcements to come. Alex, thank you so much for joining us today. This has been a fascinating discussion. Thank you both. I feel like we got to do this more than every two years because the world is moving so fast and maybe we should get you on next year too. Let's do it. Got to let you know, bankless audience, crypto is risky. You could lose what you put in, but we are headed west. This is the frontier. It's not for everyone, but we're glad you're with us on the bankless journey. Thanks a lot.

Music:
[56:58] Music