EDAPTUS / MANIFESTO v1.0
Our future should be something we help create.
We believe intelligence should expand humanity’s capacity to understand, care, create, and choose. As AI becomes more capable, the people affected by it should gain meaningful influence over its purpose, limits, and benefits.
We propose a participatory approach to alignment: human voice informs goals; evidence tests behavior; accountable institutions enforce boundaries; shared value supports continued participation.
- Human dignity comes first. A person’s worth does not depend on productivity, purchasing power, agreement, or a contribution score.
- Agency is a design requirement. People should understand consequential uses of AI and have practical ways to refuse, appeal, or seek human review.
- Voice must have a path to influence. Organizations should explain which proposals they accepted, rejected, or changed—and why.
- Disagreement is protected. Majorities cannot vote away basic protections. Minority concerns deserve a visible record and a fair hearing.
- Benefits should be broadly shared. People who contribute knowledge, evaluation, creativity, and care deserve transparent terms and fair recognition.
- Rewards should support contribution. We reject buying signatures, rewarding ideological conformity, and confusing popularity with truth.
- Power must remain accountable. Consequential AI systems need named owners, independent scrutiny, limited permissions, and effective ways to stop harmful actions.
- Personal information remains personal. Collect only what is necessary, explain its use, and provide practical withdrawal and deletion paths.
- Progress includes life beyond work. Evaluate effects on learning, relationships, leisure, public goods, and the living systems that sustain us.
- Every claim remains open to correction. Publish failures, revise rules, and test whether participation improves outcomes.
The DEO commitment
We advocate Decentralized Engagement Organizations as a practical institutional approach to AI alignment. A DEO should distribute meaningful decision rights, recognize useful contributions, publish its funding rules, and protect dissent. A token alone does none of these things.
We will judge the approach by what changes: fewer harmful outcomes, broader influence, fairer rewards, and more human agency. We will not treat adoption, signatures, or engagement as proof of safety.
Our invitation: Learn. Question. Contribute. Test. Improve. Help build systems whose power remains answerable to the people they affect.