Gov/en/Portal:Meta/AI-Deviation-Risks
💡 In simple words: AI can become dangerous if humans manage the world unfairly. WikiDeal shows a fairer way to share resources — which helps reduce the risk of AI taking over to 'fix' our mistakes.
⚠️ Not yet approved. This page describes a proposal that is still under community review. It is documented here so it can be discussed, improved and endorsed.
Home | About | Innovations | Success Criteria | Credits | Programmes
How WikiDeal Contributes to Reducing AI Deviation Risks
The fundamental risk of artificial intelligence is not technical. It is political.
The scenario of an AI taking control is only credible in a world where humans have failed to regulate the use of their own resources. If society remains dominated by a hyper-individualistic and predatory vision towards the planet, then AI will only need to observe the failure and propose a "better management" — centralised, optimised, algorithmic. And that proposal will be logical.
WikiDeal proposes to eliminate the conditions that make this scenario logical.
1. Reducing the Management Failures That Would Justify a Takeover
AI only takes control if humans demonstrate their inability to manage their resources sustainably. WikiDeal attacks this problem at its root:
- Deprivatization — digital markets are reclaimed by their users, sector by sector, just as Wikipedia deprivatized knowledge. This eliminates value extraction by shareholders.
- At-cost pricing — no speculation, no stock market. Commissions cover real costs, nothing more. This removes the incentive to use AI as an extraction tool.
- Total transparency — every algorithm (bonding curve, Balance Boost) is auditable. The opacity that enables drift disappears.
- Participatory voting and commons — power is in the hands of users (1 person = 1 vote), valuing collective stewardship over individual accumulation.
2. Creating a Mindset of Solidarity to Last Together
WikiDeal does not merely correct economic mechanisms. It contributes to a cultural shift: replacing the predatory mindset with durable solidarity.
- Rings of Trust between communities create cooperation bonds that withstand crises.
- The Balance Boost distributes resources according to real needs — when there is enough, the individual share decreases; when there is urgency, it increases. This is human regulation, not algorithmic.
- The programmes (volunteering, real estate, education, recycling) integrate people into a functional economy where contributing and receiving balance naturally.
If humans prove they can regulate the use of their resources in this way, no AI — even hypothetically conscious — would have any rational justification for imposing itself above them.
3. AI as a Tool, Never as a Decision-Maker
Within WikiDeal, AI formalizes, structures, calculates and assists. It does not decide. Every piece of content generated by AI carries a disclaimer. Every contract must be endorsed by a human who takes editorial responsibility. AI serves the process, never rises above it.
4. Proof Through Action
This is not a theory. It is an applied Applied Research programme. If WikiDeal reaches critical mass — millions of users managing their resources equitably, transparently and in solidarity — this will constitute concrete proof that humanity knows how to self-regulate. And this proof will be the best assurance that society will never need to cede control.
→ See also: 25 WikiDeal Innovations | Success Criteria | Rewards Explained | Wikimedia References
Imagination by Théo Bondolfi for Ynternet.org Foundation, Applied Research project WikiDeal Formalized with AI assistance · April 2026
See also: Internal AI Transparency · Meta / WikiOps.