Alex Blanya, co-founder and CEO of World, joins A16Z co-founder and general partner Ben Horowitz to discuss the escalating challenge of distinguishing humans from AI agents online. World is building what they call the largest real human network - a proof of human layer designed for the AI era using iris biometric verification through custom hardware devices called orbs.

The conversation explores how AI capabilities have evolved from theoretical concerns to immediate threats across social media, dating apps, video conferencing, gaming, and democratic processes. With 18 million verified users, World is deploying a global network of biometric verification devices while solving complex privacy challenges through multi-party computation and zero-knowledge proofs.

The discussion covers the technical challenges of proving uniqueness at scale, the economic implications for creator platforms and advertising, and the broader infrastructure needs for maintaining democratic institutions in an AI-dominated world.

The Three Categories of Online Identity Verification

Proof of human distinguishes between three types of online entities: humans, agents acting on behalf of humans, and autonomous agents operating independently.

"What proof of human really means is that every individual that interacts on a platform has only one, ideally one account or a limited number of accounts, and stays the owner of that account" - Alex defining the core challenge.

The hardest technical problem is uniqueness verification - ensuring one person cannot create multiple accounts while maintaining privacy and anonymity.

Why Traditional Verification Methods Fail Against AI

Web of trust systems become vulnerable when "an AI will be able to have a GitHub account and will be able to post and own an account and also attest to five other AIs that these are in fact humans, even though they're not" - Alex.

Government ID systems fail because they weren't designed for global internet platforms and create privacy concerns while lacking universal coverage.

Face ID works for one-to-one authentication but proof of human requires one-to-n verification against all previous users, creating an exponential mathematical problem.

"You can just do the math and you can calculate how much mathematical entropy, like how much information, just information theoretically, do you need to prove that" - faces and fingerprints hit walls after tens of millions of users.

Iris Biometrics and Privacy-Preserving Architecture

Iris biometrics provide sufficient entropy for global-scale uniqueness verification, with the added benefit that "Iris will turn out to be supernormal as a modality" as AR/VR adoption increases.

Multi-party computation splits iris codes into pieces sent to multiple computers so "no one actually has the information about you" and "during the computation, no one has the whole thing."

Zero-knowledge proofs separate user identity from the verification system, allowing users to prove uniqueness to platforms without World or the platform knowing anything about them.

The orb uses multiple sensors across the electromagnetic spectrum to prevent replay attacks and display spoofing during verification.

Platform Applications Beyond Social Media

Dating platforms like Tinder are implementing World verification in Japan to ensure users interact with real humans and authentic profiles.

Video conferencing faces imminent threats as "in a year from now, it's just going to be a full commodity and it's going to be super photorealistic and absolutely real time" deepfakes.

Gaming communities need protection against superhuman AI opponents, especially in competitive environments with monetary stakes.

YouTube and creator economy platforms face existential challenges when one person can "create, I think, like, it on the order of 100 videos a day on YouTube and made tens of thousands of dollars a month" using AI.

AI Manipulation Capabilities and Scale Projections

"AIs are really good at programming humans. Much better than humans are at programming AIs" - AI agents demonstrate superhuman persuasion abilities in controlled studies.

University of Zurich research on the Change My Mind subreddit showed AIs were "superhuman in their ability to change it because they were going back to the profile of the people posting it."

"What we currently see is less than 1% of what it will look like in probably a year or two" as the cost of intelligence drops exponentially and agent capabilities increase superlinearly.

Business Model and US Market Strategy

World has verified 18 million users with 40 million total in the app, but "90% of the effort of the company is just going to go about the US" over the next year.

The business requires solving three simultaneous challenges: platform adoption, device distribution (targeting 50,000 devices for sub-15-minute access across the US), and user utility integration.

"Orb on demand" service will launch soon, putting verification devices "on a motorbike and drive it to you" in major cities like the Bay Area and New York.

Distribution strategy includes large-scale partnerships with retailers like Walmart or Starbucks, individual coffee shops, and government facilities like DMVs.

Democratic Infrastructure and Government Applications

"I don't think in an AI world where you can have like very high-scale impersonation that and then with a broken social security system that like you're going to have the will of the people anymore" - Ben on democratic threats.

COVID stimulus fraud totaled approximately $400 billion, demonstrating the need for cryptographically strong citizen identification systems.

"You can buy social security numbers on the black market. Like, for those of you who don't know, that's an easy thing, that's a real thing" - highlighting current system vulnerabilities.

Mail-in ballot systems and current voting infrastructure are "built for a whole very different world" and inadequate for AI-scale impersonation threats.

Resources Mentioned

The Astrology of You and Me How to Understand and Improve Every Relationship in Your Life

connection to human, right? So you can create a pretty good podcast. Like you can take a scientific paper and give it to Gemini and say, make this into a podcast. And, you know, it'll be like a prett

I Think I Love You (a humorous romantic mystery)

e perfectly able to understand you and talk in the right way to you. For example, there's this one paper that I think you could read it after, but the Change My Mind subreddit, where the University o