On July 10, 2024, the World Foundation did something rare in crypto: they admitted their growth model was unsustainable. Phase 3 of their roadmap isn't just another upgrade—it's a pivot. For two years, World (formerly Worldcoin) paid people in tokens to scan their irises. Now, they're asking enterprises and AI agents to pay for the privilege of verifying that a user is human. The Orb will become a cash register, not a candy dispenser. This is the moment when the mask of 'incentive alignment' drops, and 'selling shovels to the AI gold rush' begins.
World was conceived as a proof-of-human protocol. Using a hardware device called the Orb, it scans human irises to create a unique digital identity, resistant to Sybil attacks. The network launched on an OP Stack L2, with a native token WLD. Phases 1 and 2 focused on distributing tokens to early adopters, building a user base of millions. The cost was borne by token inflation. But as Phase 2 ends, the model changes. Phase 3's core: sell verification as a service to enterprises, apps, and AI agents. The token incentives that drove growth are being phased out. The project that once promised universal basic income is now focused on B2B sales to the very companies that might automate human jobs away. The irony is not lost.
Let's trace the echo of trust back to its source code. World's Phase 3 is a structural audit of crypto's oldest sin: paying for adoption. For two years, World burned tokens to acquire users. Now the bill comes due. The revenue model? Verification API calls. But here's the catch: they haven't announced how payments will be made. If they accept fiat or stablecoins, the WLD token becomes a ghost—still traded but irrelevant to the core business. If they require WLD, then token holders finally get a claim on cash flows. But even in the best case, the revenue potential is tiny compared to the token's fully diluted valuation. World has ~10 million verified humans. If each verification call costs $0.01, and each user gets verified once a month, that's $1.2M annual revenue. Compare to a FDV of ~$20B. The yield is not a number; it is a narrative of risk. The market is paying not for today's revenue, but for the fantasy that World becomes the de facto human verifier for all AI interactions. That's a long bet.
We minted ghosts, but we lived in the machine. The ghosts are the millions of users who signed up for tokens. The machine is the AI agent economy. World's pivot is a bet that AI will create a massive demand for human verification. But is the Orb the right tool? AI agents need cheap, fast, privacy-preserving verification. Orb scanning is expensive (hardware cost, user friction), and faces privacy backlash. Cheaper alternatives exist: behavior analysis, zero-knowledge proofs of social connections, even captcha 2.0. World's hardware moat is also an anchor. Let's not ignore the regulatory elephant: GDPR investigations, ICO complaints. Phase 3 doesn't solve that. If regulators ban biometric data collection, the Orbs become paperweights.
Based on my audit experience with identity projects, the gap between stated mission and actual code is widest when the mission shifts. I remember auditing ICO whitepapers in 2017, looking for the gap between promise and code. This feels eerily similar. World's GitHub shows ongoing development on the L2, but no public API for verification services. Truth hides in the silence between the blocks. Until we see a smart contract that accepts WLD for verification, this is pure marketing.
Now let's examine the AI agent market. Gartner predicts that by 2027, 40% of online interactions will be mediated by AI agents. These agents will need to distinguish humans from bots for tasks like voting, customer support, content moderation. World is betting that a physical hardware-based biometric proof will be the gold standard. But AI agents don't have eyes. They consume metadata, not iris scans. A simpler solution: a signed attestation from a social graph or a device attestation (like Apple's Secure Enclave) could suffice. World's Orb adds friction to the user journey. The question is whether enterprises will pay a premium for stronger guarantees. So far, no enterprise contract has been announced. The market is pricing hope, not evidence.
The contrarian angle: Phase 3 is not a pivot to revenue; it's an elegant exit strategy. By framing the end of token incentives as 'Phase 3', World avoids the narrative of failure. They can claim they've 'succeeded' in reaching enough users and now need to monetize. But what if the real story is that token incentives were unsustainable, and the market had priced in the token's eventual collapse? By shifting to service revenue, they create a new plotline that may never deliver, but keeps the narrative alive long enough for insiders to exit. Meanwhile, the WLD token continues to trade, buoyed by hope. The ghosts of ICOs past whisper: this is the same Siren song, just dressed in AI makeup.
Moreover, consider the competitive landscape. ENS offers a decentralized identity without hardware. Polygon ID uses zero-knowledge proofs on mobile. Gitcoin Passport aggregates social credentials. All are cheaper and less invasive. World's only edge is the hardware-based uniqueness, but that comes at a cost: each Orb costs roughly $1,000 to manufacture. To reach 100 million users, they'd need 100,000 Orbs, a $100 million hardware bill. The token inflation covered this before. Now, service revenue must cover it. At $0.01 per verification, they'd need 10 billion verifications to break even on hardware alone. That's 1,000 verifications per user over their lifetime—unlikely unless verification is needed daily.
There's also the ethical dimension. World's original pitch was about universal basic income and financial inclusion. Phase 3 quietly drops that for profit. The mission drifts from 'every human gets a share' to 'every AI agent pays a fee'. The same INFJ instinct that made me write about ICO illusions now sees a new illusion: that you can build a trusted identity network by scanning the most private part of the human body—the iris—and then sell access to that data to the very machines that threaten human agency. We minted ghosts, but we lived in the machine.
The takeaway is not a conclusion, but a question. The question isn't whether World can sell verification. It's whether the world needs a machine that scans your soul to confirm you're human. AI agents might not care about your iris—they might just ask for a cryptographic signature from a device you already own. Phase 3 is a gamble that the future of trust is centralized hardware, not decentralized software. I'd rather watch from the sidelines, waiting to see if the silence between the blocks reveals a truth or a tomb.
For traders, this is a narrative play with binary risk: either World signs a major AI deal (moon) or regulatory action kills the Orb (crash). The current price reflects neither extreme, but the volatility will spike. For long-term believers, wait for three signals: first, a published API accepting WLD; second, a named enterprise partner; third, a revenue report in the monthly update. Until then, the yield is a story, not a number. And I have learned to be skeptical of stories that end with 'Phase 3'.