Aiki Wiki

The Win Win Protocol for the World Wide Web

Aiki Wiki is a computational system for online conversation and consensus building that facilitates a mutually rewarding outcome and a perfect accounting for all possible events in the chaos of human conversation on the web, everything from brainstorming to conflict resolution is transmuted into a “win-win” conversational sport.

Aiki Wiki is a type of conflict resolution and brainstorming algorithm that can help filter both social media as well as AI systems for alignment, perfectly. Aiki Wiki can become the first exportable protocol that offers platforms and systems highly competent tools for online disinformation, misinformation, harassment, reliable fact checking, in addition to a sophisticated dispute and conflict resolution tool.

The “Win Win Protocol” applies something like Game Theory to conversation which facilitates thorough, mutually satisfactory governance transactions or decision making without voting, without any third-party mediation, without censorship nor requiring any advanced AI systems like LLMs, just by having or processing conversations through its protocol.

The protocol has endorsements from law, computer science and cognitive sciences.

Improve everything, replace nothing

The system’s emphasis on creating a psychologically safe environment where users can acknowledge mistakes or uncertainties encourages honesty and rationality. LLMs can be trained to create similar environments by being transparent about their limitations and aligning their responses with users’ needs for trust and safety, fostering alignment with user values.

The system’s focus on resolving conflicts through a structured narrative provides a model for LLMs to mediate disagreements or differences in user input. By learning from the system’s nine narrative events that guide conversations from conflict to resolution, LLMs can be trained to facilitate constructive dialogues, helping align their outputs with user intentions and fostering productive discussions.

The “Non-Dual Consensus Theory” provides a robust framework for aligning LLMs with human values through collaborative decision-making, handling ambiguity, resolving conflicts, personalizing guidance, fostering trust, and using heuristics. By integrating these principles, LLMs can better align with user expectations and societal norms, leading to more harmonious and beneficial interactions.

Let’s clean up social media, create true “town squares”, and align AI systems in a trusted “win-win” outcome.

Let’s move quickly and fix things.

This project is proudly funded via “Bootstrapping Arts and Engineering”