Sybil and the Machine: Myth, Multiplicity, and the Crisis of Verification

Sybil and the Machine: Myth, Multiplicity, and the Crisis of Verification

By Sasha Shilina

“I myself saw the Sibyl at Cumae with my own eyes, hanging in a jar; and when the boys asked her, ‘Sibyl, what do you want?’ she answered, ‘I want to die.’”
– Petronius, Satyricon, 48.

I. The problem of too many voices

Web3 was built to welcome the many, and then discovered it couldn’t tell who the “many” were. Wallets are cheap, scripts are tireless, and incentives attract shadows. We gave this failure of recognition a mythic name: the Sybil. In cryptography, a Sybil attack is one actor pretending to be many to skew coordination and drain public goods.

But the myth is deeper than the term. The Sibyl of antiquity spoke in fragments, leaves scattered by the wind, truth in multiplicity. Today’s networks hear multiplicity and panic. 

How do we keep the network open to many voices without being fooled by one ventriloquist?

II. The myth inside the protocol

Humanode’s wager is simple and radical: anchor voice in a living, unique person. Privacy-preserving liveness checks make “one person = one node = one vote” feasible without revealing identity or storing raw biometrics [1]. It’s a counter-myth: against the many masks of Sybil, the protocol asks for a single, embodied presence.

This solves fraud. It also risks flattening. If “real” means “singular,” what happens to the parts of self that are contextual, protective, playful, or artistic? The danger of making visibility the price of participation is old news in political theory: Michel Foucault warned that “visibility is a trap” [2].

III. Identity is plural—by design

Offline and online, selves are distributed. We are different at work and at home, in an art DAO and a research DAO, under a legal name and a pen name. Some multiplicity is safety. Some is play. Some is care. Canadian sociologist Erving Goffman called it the presentation of self across stages, while Judith Butler showed how identity itself is a performed repetition that never fully settles [3,4].

Systems that equate personhood with one uniform persona confuse uniqueness with sameness. They succeed at verification and fail at recognition. Helen Nissenbaum’s contextual integrity is instructive here: privacy and identity make sense only within a context, not across all contexts at once [5]. Édouard Glissant goes further and defends a right to opacity, the ethical space not to be fully legible to power or platform [6].

  • Verification asks: Are you alive and singular?
  • Recognition asks: Which self is speaking, to whom, under what risk, with what responsibility?

We need both.

IV. Ethics of liveness

Liveness is an ontological minimum: proof of zoē (bare life), not bios (a life narrated and situated) [7]. It protects against bots; it does not honor the human need for opacity, context, and role-shifting. When liveness becomes the whole of personhood, it can quietly punish those whose bodies or circumstances sit outside normative inputs, and those who need many faces to stay whole. A scholar of race and surveillance Simone Browne warns that biometric regimes often reproduce existing exclusions; anthropologist James C. Scott shows how systems that demand legibility can misread and remake the world they claim to see [8,9].

V. Protocols of plurality (design moves)

  1. Rooted PseudonymyOne biometric root → many contextual sub-identities (personas) cryptographically linked to the root but unlinkable to each other by default. Governance can require “one-persona-per-context” without collapsing all contexts (selective disclosure via SSI/VCs).
  2. Scoped UniquenessEnforce uniqueness within a domain (one human—one vote in DAO X), not globally. Prevents cross-domain deanonymization while preserving anti-Sybil guarantees where it matters.
  3. Plural ReputationKeep per-persona reputations instead of a global score. Allow users to carry only what they choose across contexts using zero-knowledge proofs (e.g., “has contributed ≥N times” without revealing identity).
  4. Consent-Aware LinkingAllow voluntary, one-time proofs that two personas belong to the same human for conflict-of-interest checks, then cryptographically forget the link (burn-after-verification attestations).
  5. Graceful Failure & AppealsAlternative verification paths (human review, assisted capture, multi-modal signals) when liveness fails for disability, trauma, environmental or device limits. Personhood shouldn’t hinge on a camera’s mood.
  6. Deliberative ModalityDifferent decisions, different thresholds. Use one-person-one-vote for constitutional choices; allow contribution-/expertise-weighted voting for operational matters, with personhood as the non-negotiable lower bound.

VI. Reframing Sybil: from enemy to teacher

The original Sibyl’s fragments weren’t fraud; they were polyphony. The lesson isn’t that multiplicity is bad, but that our interpretation is brittle. We can build systems that tell a choir from a ventriloquist. Think of Humanode not as a demand to be one thing forever, but as a root of trust from which many responsible selves can grow.

VII. Conclusion: listen without losing the signal

A future-proof public sphere needs two virtues at once:

  • Clarity—each vote, one living person.
  • Mercy—room for masks, roles, transitions, and silence.

If we only verify the face, we’ll miss the person. If we only honor the voice, we’ll lose the commons. The work ahead is to bind presence without binding identity, to keep the system honest and the self alive to its own plurality.

The Sybil was never just an attacker. She was a warning and a guide.

Build for the many—without letting one become many in bad faith.Build for the one—without amputating the many within.


Footnotes

  1. Humanode Core Team, Humanode Whitepaper v2.1 (2025), esp. proof-of-biometric-uniqueness, privacy, and liveness sections. https://whitepaper.humanode.io
  2. Michel Foucault, Discipline and Punish: The Birth of the Prison (1975), “Panopticism”: “Visibility is a trap.” 
  3. Erving Goffman, The Presentation of Self in Everyday Life (1956).
  4. Judith Butler, Gender Trouble (1990), esp. “gender as the repeated stylization of the body.” 
  5. Helen Nissenbaum, Privacy in Context: Technology, Policy, and the Integrity of Social Life (2010). See also her 2004 paper on contextual integrity.
  6. Édouard Glissant, Poetics of Relation (1990), on the right to opacity.
  7. Giorgio Agamben, Homo Sacer: Sovereign Power and Bare Life (1998), distinction of zoē and bios.
  8. Simone Browne, Dark Matters: On the Surveillance of Blackness (2015), on biometrics and exclusion.
  9. James C. Scott, Seeing Like a State (1998), on legibility and misrecognition.