Technical Whitepaper
Version 1.1 Draft
The HUMAN Protocol is a broadly applicable approach to organizing, evaluating, and compensating human labor. It is designed to enable a new generation of machine intelligence to apply human labor to self-improvement in order to achieve human parity in task performance.
Today this work is commissioned by machine learning practitioners. The protocol’s immediate application is thus to improve the most labor intensive problems in machine learning: making datasets fit for training via annotation and validating model inference quality.
While the HUMAN Protocol supports and improves today’s practices, it is engineered for the next evolution of human inputs to machine intelligence: letting machines ask people directly for the data they need to improve.
HUMAN Tokens (“HMTs”) serve as the medium of exchange in the HUMAN Protocol. They are EIP20-compatible tokens, and the complete system forms a decentralized platform with an open protocol. Each component receives a fee for its role, and interactions are coordinated via smart bounties on the Ethereum blockchain.
MVP reference implementation is complete, and the HUMAN Protocol’s first app (hCaptcha.com) is live on the Ropsten testnet, serving many billions of requests per month. Development has been underway since 2017, and open source implementations of many protocol components are already becoming available for different use cases. For example, Intel’s open source CVAT tool was ported to run on the HUMAN stack in December 2020.
Background: the hCaptcha.com app provides a dropin replacement for reCAPTCHA, and demonstrates the scalability of the underlying HUMAN technology. It equally provides a platform to displace Mechanical Turk, Figure Eight, and other intermediaries for some types of tasks. Requesters compensate website owners (“miners”) based on challenges answered and computed accuracy. HMT mining power is thus directly proportional to human labor.
The core ideas of the HUMAN Protocol are broadly applicable and quite abstract, so we have chosen to write this overview with frequent examples from a particular application (hCaptcha.com) to demonstrate each component’s role within that context.
Last modified 6mo ago