[Intermediate] — ENS, the Sybil problem, and why decentralised identity is the field’s most important unsolved problem.
Why this matters more than it looks
Almost every hard problem in this folder reduces to not being able to tell people apart:
- Governance is plutocratic because one-person-one-vote requires knowing what a person is
- Lending must be over-collateralised because anonymous default is free
- Airdrops are farmed by thousands of wallets belonging to one person
- Spam and Sybil attacks are unpreventable because identities are free
Blockchains deliberately made identity free and unlimited — that’s what “permissionless” means — and then inherited every problem that follows. Solving it without reintroducing a central authority is genuinely hard, and it’s the field’s most consequential open problem.
ENS — the clear success
The Ethereum Name Service maps human-readable names to addresses and much else:
vitalik.eth → 0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045
→ a content hash (an IPFS site)
→ an avatar, a text record, other chains' addresses
Why it works when so much else hasn’t:
- It solves a real, felt problem. Nobody wants to paste a 42-character hex string, and sending to a wrong address is irreversible
- The token is the asset. An ENS name is an NFT, and there’s no off-chain thing it points at that could disappear → NFTs in practice
- Renewal fees prevent permanent squatting and fund the protocol
- It’s a general resolver, not just addresses — content hashes let you host a censorship-resistant frontend at
yourapp.eth→ storage
Support ENS in any dapp you build. Resolve names in inputs and display them instead of addresses — it’s a small change that measurably reduces the most costly user error there is.
Others exist (Unstoppable Domains, Solana’s SNS, Farcaster names). ENS is the one with real integration.
The Sybil problem
Creating an address is free and unlimited, so any system distributing anything per-address is farmed:
- Airdrops — professional farmers run thousands of wallets, each performing the qualifying actions. A large fraction of most airdrops goes to farmers, not users
- Governance — splitting tokens across wallets is trivial (though token-weighting means it doesn’t help, which is why plutocracy is the stable equilibrium)
- Quadratic funding — collapses entirely without Sybil resistance, since it weights by contributor count
You cannot solve this with cleverness on-chain. Something outside the chain must attest that an identity is distinct — and that something is a trusted party, which is the thing being avoided.
The approaches, and their honest status
Proof of personhood.
Worldcoin scans irises to produce a unique biometric hash. Technically ambitious, with zero-knowledge proofs so the biometric isn’t stored. Banned or investigated in multiple jurisdictions (Spain, Portugal, Kenya, Hong Kong) over biometric data collection, and the practice of gathering iris scans from people in low-income countries in exchange for tokens drew heavy criticism. The clearest illustration of the trade: to prove uniqueness you must collect something unique about a person, and that is exactly what privacy law restricts.
Proof of Humanity / BrightID use social graphs and video verification. Less invasive, much smaller scale, and harder to grow.
Verifiable credentials and DIDs. W3C standards for issuer-signed claims you hold and present selectively, ideally with a zk proof so you reveal “over 18” rather than a birthdate. Technically sound, standards-mature, and adoption remains limited — the bottleneck is that issuers (governments, universities, employers) must participate, and they mostly haven’t.
Soulbound tokens. Non-transferable tokens representing credentials, attendance or membership. The right primitive for credentials, since transferability defeats the purpose. ERC-5192 standardises them. Modest adoption, and no answer to the underlying uniqueness problem — a person can still hold many wallets.
Attestations. The Ethereum Attestation Service lets anyone sign a claim about anything. A useful primitive rather than a solution — it moves the problem to “whose attestations do you trust,” which is a social question, not a technical one.
On-chain reputation. Score an address by its history — transactions, repayments, longevity. Cannot distinguish “one honest person” from “one of a farmer’s thousand wallets,” but it does raise the cost of farming, and it’s the basis of the tentative work on under-collateralised lending.
The unavoidable tension
Every approach that actually works reintroduces a trusted party or collects personal data. That’s not a failure of imagination — it’s structural:
Proving uniqueness requires something unique to a person, and something must vouch that it is.
The genuine progress is in selective disclosure: zero-knowledge proofs let you prove a credential’s implication without revealing the credential — “I hold a passport from an EU country” without which country, which passport, or which person → zk proofs. That’s real, and it narrows the trade rather than eliminating it: you still trust the issuer, but you no longer expose the data to every verifier.
What’s usable today
- ENS for naming. Mature, use it
- Sign-In With Ethereum (EIP-4361) for authentication. Standardised and simple → wallets
- Attestations for application-specific claims where you choose the trusted issuers
- Soulbound tokens for credentials
- Nothing for general-purpose Sybil resistance. If your design needs it, design around it instead — cap per-address impact, use time-weighting, or accept a trusted issuer explicitly
Key insight
Permissionless means identities are free, and free identities break every one-per-person mechanism. This is a design consequence, not an oversight — and it’s why governance is plutocratic, lending is over-collateralised and airdrops are farmed. The field’s most valuable open problem is proving uniqueness without a central authority, and every partial answer so far trades away either privacy or decentralisation.
Related
- DAOs and governance — where this bites hardest
- zero-knowledge proofs — selective disclosure
- wallets and connection — SIWE
- secure authentication — the ordinary version
Software engineering
Infrastructure & security
Computer science
Hardware & engineering