Human–AI collaboration is the externalisation of human intentions into a non-human system that can act on it as either a supplement to, or replacement of, human activity. These collaborations also enable results impossible to achieve by humans alone, representing a new category of transhuman work. Protocols are the mechanism for initiating, refining, coordinating, monitoring, validating, and completing transhuman work.
Effective human collaboration methodologies share design properties — data gathering, clarification practices, inspectability and monitoring (including failure handling), accountability (such as roles, domains of influence, and trust calibration), treatment of shared context at multiple horizons, and standards for language, tooling, and behaviour. These recur across traditions over time, including modern paradigms designed for knowledge work (GTD, agile, pair programming).
While human methodologies manifest as habits, rituals, and ceremonies, transhuman collaborations require protocols. It is our hypothesis that human collaboration paradigms can be transposed into transhuman protocols by deriving fundamental characteristics from concrete implementations of human paradigms, and recasting them for transhuman use.
We also intend to explore protocols for inter-human communication among diverse neurotypes as well as inter-species communication as avenues for productive borrowing and/or contrast with mainstream inter-human protocols.
This research aims to characterise these properties using novel and existing protocol description paradigms to produce a working library of protocols.
We further intend to measure these protocols against a stated definition of effectiveness. This definition should involve not only comparing outputs against the stated (and evolving) intent, but wider contextual implications for individuals (such as cognitive load / sovereignty / debt) and groups (governance, bias, and social impact). In this way it can serve as a holistic measure of successfully applying transhuman practice within existing systems of human collaboration.
The library and its findings is intended primarily for an audience of protocol and systems researchers as a basis for further investigation, but should be of equal interest to those studying and facilitating transitions into transhuman ways of working.
Updated 10 May 2026 by Robert Peake