The Limit of Personhood / Proof of Personhood

The proxy problem

No system verifies personhood. Systems verify proxies for personhood — possessions and traits that we hope correlate with being one unique human:

Something you have: a phone number, a government ID, a hardware key.
Something you are: a face, an iris, a fingerprint, a voice.
Something you did: a history of behavior expensive to fake.
Something you stake: money or reputation you would lose by cheating.

Each proxy is a bet that the correlation holds. The history of the field is the history of those bets failing.

Three ways every proxy breaks

A proxy can fail in three directions, and most fail in more than one.

Forgeability. If the proxy can be manufactured or rented more cheaply than the value of passing, attackers will. SIM farms, deepfake liveness, purchased credentials — the supply chain for fake humans is mature and global.

Exclusion. Tighten the proxy to resist forgery and you start rejecting real people. Not everyone has a government ID, a smartphone, a bank account, or a face a camera reads correctly. Every increase in rigor is paid for by someone who is exactly who they say they are and gets turned away.

Surveillance. Make the proxy strong and persistent and you build a system that knows, permanently, which human did what. Solving impersonation by eliminating anonymity is not a solution. It is a different catastrophe.

A perfect scheme would be unforgeable, universal, and privacy-preserving at once. No scheme is all three. You choose which failure you can live with.

Uniqueness is harder than existence

Proving a person exists is comparatively easy. Proving a person is unique — that this human has not already enrolled under another identity — is the hard core. This is the Sybil problem: one person, many masks.

Uniqueness requires a global view. To know you have not registered twice, the system must compare you against everyone. But comparing biometrics across everyone is precisely the surveillance apparatus we were trying to avoid. Uniqueness and privacy pull in opposite directions, and the tension does not resolve; it only gets relocated.

The definitional ceiling

Here is the limit. Suppose every technical failure were fixed: an unforgeable, universal, private proof. We would still face the question the technology cannot answer — what is a person?

Is a person a body? Then what of two people sharing a degraded body, or one mind across two? Is it a continuous biography? Then identity becomes memory, and memory can be implanted or lost. Is it legal standing? Then personhood is whatever a state grants, and the state can revoke it.

The machine needs a crisp predicate. Personhood is not crisp. We are asking computation to draw a hard boundary around a concept that human beings have never managed to define, across millennia of trying.

Living at the limit

The honest position is not despair but calibration. We will never prove personhood; we will only raise the cost of faking it and lower the cost of being recognized. The goal is not a perfect gate but a good-enough asymmetry: cheap for a real person to pass once, expensive for an attacker to pass a million times.

That reframing matters. It moves the field from truth to economics — from certifying souls to pricing fraud. We are not building a detector for personhood. We are building a tollbooth that real people can afford and fakes cannot.

The limit of personhood is that the toll can never become certainty. It can only become enough.