Led by the 10-year-old venture firm Blockchain Capital, whose bets have included Coinbase, Kraken and OpenSea, the investment brings Worldcoin’s funding to at least $240 million, even as the controversial organization — founded in 2019 by OpenAI CEO Sam Altman — has much to prove.
Yesterday, we talked with Blockchain Capital General Partner Spencer Bogart about what gave him confidence in Worldcoin, which aims to create a global ID, a global currency, and an app that enables payment, purchases and transfers. Like many others, we wondered how it can achieve its goals when, right now at least, its mission relies first and foremost on convincing tens of millions of people to allow Worldcoin to scan their irises using futuristic, tech-dense globes.
Below is part of that conversation, edited for length. You can also hear the longer conversation here.
Your co-investors in this new round include earlier backer Andreessen Horowitz, Bain Capital Crypto and Distributed Global. Did Khosla Ventures or Tiger Global, which are also previous backers, re-up?
They might be part of this financing; I don’t believe they’re a large part of it.
How much of the company do investors own? I’d guess it’s hard to negotiate with Sam Altman given the power he wields and also his extensive experience on the other side of the table as an investor.
That is a correct characterization. Sam is a formidable founder and knows how to manage a cap table. Again, I apologize. It’s not a figure I have in front of me right now. Generally, companies sell 20% of the [equity] in each financing. Granted, things can move down or up from there significantly. I think in this case, the number is going to be meaningfully lower than that across the Series A, the Series B, and the Series C.
How long had you been talking to Worldcoin, and what motivated you to lead this deal?
The original genesis was Sam wondering: what if I could create a cryptocurrency that I could distribute to everyone in the world and everyone got an equal share of it? For me, from a venture perspective, that’s certainly interesting [though] I don’t know that it’s something that we would be super excited to go and underwrite based on the things that our team is typically interested in.
[Meanwhile] this requires basically making sure that no one person can accumulate a disproportionate share of it, which requires people to be able to identify unique humans. And this gets into really the part that we’re excited about, which is World ID. It’s this ability to easily distinguish between machines and humans on the internet [because] most of the internet is supported by ad revenue and it costs just as much to serve as bot traffic as it does to service human traffic. It’s why various applications and service providers have used CAPTCHAs to distinguish between bots and humans. But that’s no longer viable in a world of advanced automated systems and particularly things powered by AI. It also doesn’t differentiate between unique humans, so I don’t know if the same person is coming to consume a resource excessively
That leads us to: okay, how can we provide a means of distinguishing between humans and bots and make sure that each human is unique?
Which leads to biometrics.
The root of what defines humans is biometrics, and my first thought was: why create this custom hardware to go scan eyeballs? Like, billions of people are already walking around with an iPhone. Why don’t we use Face ID, right? The problem is that human facial structures do not have sufficient randomness or entropy to distinguish between unique humans, at the scale of tens of millions or hundreds of millions of people.
I didn’t realize that was the case.
It’s not something that occurred to me either. I didn’t think about the fact that once you get past one hundred million people, there are going to be a lot of people that look like Spencer Bogart; their facial structures are going to be sufficiently indistinguishable from mine. Fingerprints have the same problem; there’s not sufficient randomness in fingerprints.
That leads us to two viable options, DNA that does have sufficient randomness to be able to prove human uniqueness at the scale of billions of people. But you’re providing way too much information with DNA. Then there are irises. As it turns out there is an insane amount of entropy and randomness in the human iris. And in this case, the team has built an insane amount of protection. You get an iris scan, it does not store your iris by default. It is deleted on the device immediately. It is only used to create what’s called an iris code, which is a unique mapping or encoding of your iris. And it’s compared against all others. And now, with these iris codes, we don’t know their name or location or anything. The only thing we know about all of them is that they’re unique human beings.
I’m guessing an enterprise strategy — helping companies cut down on their interaction with bots — is the most lucrative opportunity right now for Worldcoin. You could also deliver this cryptocurrency to everyone, though it’s not clear to me how people would use it. But before any of this can happen, you need to get a meaningful number of people in front of these orbs that are strange and not readily accessible, when people are already nervous about biometrics and cryptocurrency. Worldcoin says it has now scanned the eyes of 2 million people. How many does it need for this to become meaningful? One billion?
These are the right questions. It’s about: do you have a network of provably unique humans? And that is only going to be interesting to applications and enterprises at a certain scale. But I think it’s going to depend on use case. By the time you get to 10 million unique users, there’s already a range of applications that would like to use that, while others are not going to be interested in using it unless you’re at a network of 500 million or a billion or 2 billion people.
Some of the other challenges here are yes, obviously, orb distribution. There are currently 200 to 300 [orbs] in the wild today, with another 2,000 that have been manufactured and are waiting to be deployed. Then there’s this question of public perception. Something that we flagged as part of the investment is: is there going to be so much negative perception of this that no matter how much we’re confident that this is 100% viable, is public perception going to be so negative that people will not want to participate?
So far, the data says otherwise. Worldcoin has already onboarded nearly 2 million people by operating a pretty capital-intensive boots-on-the-ground strategy, and this is just in beta testing. This is without pushing or pulling any levers on marketing; this is without having the protocol even live on mainnet. This is only in initial testing.
As for some of the things that might use this, Elon Musk has talked a lot about a bot problem on Twitter, and has touted the idea of if we make everybody pay $8 a month, that’ll help solve the bot problem. We think that World ID is a lower-friction way of solving the same problem and will be a higher fidelity solution. And there is a range of new applications and services that have not existed because of our inability to make this distinction historically. What those are, I don’t know, but we’re interested in funding them.
Again, you can hear much more about the investment here, including why OpenAI might itself become a major customer of Worldcoin some day, why Bogart wasn’t bothered when hackers recently installed password-stealing malware on the devices of multiple Worldcoin orb operators, and why he’s fascinated by flash trades on the blockchain.