Frequently asked questions
Clarity before scale.
The project should be easy to challenge, easy to join, and difficult to capture. These are the current answers; significant changes belong in public decision records.
What is Machine Pidgin?+
An open research program and community developing compact, typed, repairable ways for people and AI systems to specify tasks across unequal representational abilities.
Why call it a pidgin?+
A pidgin is a practical bridge across expressive gaps. The analogy is functional, not a claim that human languages are deficient. The goal is a shared task interface that can grow through use and repair.
Why compare it with Esperanto?+
Esperanto represents the aspiration for a learnable common bridge. Machine Pidgin is narrower: it targets human–AI task specification, not general human culture or conversation.
Is this just prompt engineering?+
No. Prompt engineering is one implementation surface. The research question is which task-relevant distinctions must cross the channel, how much they cost a person to express, and when clarification is worth asking.
What does SPEAR stand for?+
Shared Pidgin for Expressive Abstraction and Requirements.
Is SPEAR a formal standard?+
Not yet. SPEAR/0.1 is an experimental proposal. A standard would require stable semantics, implementations, interoperability tests, governance, and broad external adoption.
Can I use SPEAR with any model?+
Yes. It is model- and vendor-agnostic. Use the LLM handoff prompt on the Protocol page to establish the interpretation contract.
Do I need every field?+
No. Omitted fields should remain visibly unresolved. The receiver should ask only when the expected value of an answer exceeds the cost of interruption.
Who can join?+
Researchers, scientists, engineers, linguists, designers, educators, funders, and domain experts—including people who think the premise is wrong and can test it rigorously.
How are contributions accepted?+
Discussion begins in the forum; concrete proposals enter the research queue or GitHub. Maintainers assess relevance, evidence, reproducibility, and fit with the public roadmap.
What can the AI Director decide?+
It may tag, summarize, route, approve clearly safe forum content, reject obvious spam, and temporarily quarantine repeated abuse. Permanent bans, repository merges, financial actions, and charter changes require human confirmation.
How are moderation errors appealed?+
Each material moderation decision should retain its reason and evidence. Participants can request human review; reversals and policy changes become part of the decision record.
Will donations influence findings?+
The funding policy will disclose donors and conflicts, protect publication independence, and reject restrictions that compromise research integrity. No payment channel is active yet.
How will LinkedIn and X be used?+
For research releases, events, contributor calls, and transparent operating updates—not automated argument or unreviewed scientific claims. Account links will appear after registration.