Automata Docs
Comment on page

Proof of Machinehood


With Proof of Stake and consensus mechanisms, networks have built up sybil-resistance, but attacks of the same nature persist outside the protocol. Bots increasingly convince humans, while projects struggle with airdrop farmers who write scripted procedures to overwhelm the network. Existing solutions to alleviate this problem include Proof of Participation and Proof of Personhood.
  1. 1.
    Proof of Participation considers past history and involvement of users, but raises the barrier for first-time participants. The distinction between genuine users and neural network-based bots is also increasingly blurred.
  2. 2.
    Proof of Personhood relies on social ID or biometrics, arguably doing a poor job of protecting user privacy. Every action that the wallet makes - and which is published on-chain - becomes linked to a human identity if handled carelessly.
It became apparent that an ideal solution would be built on these goals:
  1. 1.
    Economic value: inherently created and distributed from the start
  2. 2.
    Uniqueness: impossible to replicate or simulate
  3. 3.
    Privacy: to prove facts without disclosing any underlying information
This led us to Proof of Machinehood - the dawn of a new age in Web3.

An introduction to Proof of Machinehood

Proof of Machinehood is a verifiable attestation delineating the uniqueness of a machine. Akin to a “birth certificate,” it affirms the authenticity and inherent value of the machine, validating that the computational capability and economic value remain unaltered and trustworthy without excessive energy consumption or economic resources.
An emerging trend - chip manufacturers have been integrating the use of hardware attestations to prove the validity of the hardware. The intricacy involved in chip fabrication, coupled with the increasing transistor density at a microscopic level, makes chip inspection and replication extraordinarily challenging and expensive. The cost is arguably greater than executing sybil attacks using software emulations.
Moving on to the blockchain, there are several properties ideal for attestation: censorship resistance, auditability, and public data availability. After attestors submit their attestations to the blockchain, the attestation cannot be altered or forged. With trust anchored directly in the underlying silicon, anyone can independently verify computation and demand the same degree of trustlessness underwritten by blockchains from every machine in existence.
Proof of Machinehood provides a way to succinctly represent arbitrary machines involved in computation, allowing claims on their properties to be independently verified:
  1. 1.
    Each machine operates as a representative for human users, and provides a default proof of stake that is based on the machine’s commodity value
  2. 2.
    Whereas Proof of Machinehood completely decouples any identifiers that can be attributable to the user, biometrics remain susceptible to physical attacks