Building the Wikimedia Cathedral: A 100-Year Vision for AI-Augmented Free Knowledge

Related Links
Video of panel discussion @ Wikimedia Future Lab

First of all, I’d like to express my gratitude for having been invited as an external expert to the Wikimedia Future Lab conference. It was an energizing and hopeful gathering and I deeply appreciated the open spirit, the readiness to confront complex challenges and the shared passion for open knowledge.

Allow me to share a synthesis of my analysis, along with an open invitation: I’d be honored to deliberate further with Wikimedia leadership, subject matter expert staff and, crucially, with community stakeholders, to co-develop the strategic scaffolding needed for the critical transition of Wikimedia into the AI era. We have the chance not just to iterate, but to co-create a vision and roadmap that stretches beyond the short-term horizon and dares to think in ‘cathedral time’.

The Case for Boldness: Wikimedia as a 100-Year Institution

Throughout the conference, I heard a recurring call to action: we must act now. This is not alarmism; it is a reasonable recognition that the next chapter of open knowledge cannot be written with yesterday’s infrastructure and mental models.

As I expressed during the panel discussion, I believe now is the moment to conceive a longterm vision for Wikimedia – a cathedral built with the open, generative principles of the bazaar. A Wikimedia that holds its foundational values, while acquiring the strategic foresight, institutional depth and strategic partnership capacity necessary to serve as an anchor for a transnational public interest knowledge ecosystem for the next century.

A bold future requires bold bets. Below, I outline three interconnected initiatives that I believe have the transformative potential to propel Wikimedia into the AI era, while remaining deeply human-centered and public interest:

1. Personalization: A Knowledge Tree for Every Human

Imagine if every user could engage with Wikipedia not just as a repository, but as a canvas for their own learning journey. A space where you curate your personal knowledge tree as living record of your intellectual development, passions, and questions.

These trees could have dimensions that are deeply private, others that are collaborative, as well as others that are a curated version, that is shared as part of one’s public identity. They would offer continuity in a fragmented information landscape and position Wikimedia as a lifelong anchor for learning and meaning-making.

Far from undermining neutrality or openness, personalization (if done ethically and transparently) reinforces Wikimedia’s commitment to relevance, resonance, and dignity of difference.

2. Learning & Education: From Codifying Knowledge to Facilitating Growth

Wikimedia’s DNA is knowledge codification. And that should remain central, but in the era of AI expanding our scope to learning becomes vital. We must IMHO evolve from documenting knowledge to cultivating capability; especially learning co-creation/peer-production as well as effectively and fairly contributing to collective decision making (i.e. governance and collective action).
This would mean extending Wikipedia into an engine for collaborative learning. I am thinking along the lines of: (1) A curriculum around collabowriting, co-creation and peer-production skills; importantly training the next generation of wikipedia contributors. (2) A platform to allows to host local and transnational wiki-learning communities that parents and educators can easily enroll students/kids into.

This is not just about outreach, it’s about developing learning infrastructure. A next-generation learning ecosystem that makes Wikimedia a foundational part of education worldwide.

3. Embracing AI: WAikiki and the Augmented Collective

Let’s stop fearing the AI behemoth and start designing for sovereign co-agency between humans and AI. I see AI integration along the following dimensions: 

3.1. Human-AI Interaction: From Commands to Intent

AI changes the paradigm of human-computer interaction. Instead of giving explicit instructions, we articulate intent. This opens the door for new UX concepts (i.e. Wikipedia can not just remain a Website). Let’s think about concierge agents that assist newcomers, guide editors through contribution flows and adapt navigation based on users’ learning goals. This has the potential to significantly lower friction and provide a welcoming, hand-holding user experience.

3.2. One Wikipedia: Shared Backbone, Adaptive Layers

Imagine one core Wikipedia set of articles structured around our knowledge graph. It would allow editing that ONE core of articles collectively while delivering articles in all languages, with localized and personalized customization tailored to each user’s context, background and intent. Done right, this would enhance quality and accessibility across linguistic communities.

3.3. WAiki: Wikipedia + AI as the Public Interest Companion

Another exciting prospect set of value propositions of WAiki as framework for AI-powered augmentation of Wikipedia as a public infrastructure would be:

• An omnipresent personal fact-checker (for all media and information received)
  
• A learning engine and knowledge navigator that scaffolds each user’s lifelong inquiry
  
• An identity- (or reputation-) anchor/graph that reflects one’s interests (ideally also badges representing competences) as well as contributions (to private- and public-interest enterprises)
  
• Together the WAiki functionalities can serve as a social layer that nourishes belonging and coordination for meaningful collective action.
  

With these affordances, Wikimedia will become one of the most potent testbeds for human-AI collaboration at scale, exploring new frontiers of collective intelligence.

Conclusion: The Time Is Now

The urgent challenge and opportunity before us is not just technical and economic, it is deeply cultural and it requires us as community to embrace transformation (and give up or evolve habits and pillars of our collective identity). Wikimedia can become evolve as sanctuary for open knowledge, a beacon of Enlightened education and participatory democratic governance and hence a prototype for how humans learn and build together in the AI era.

Let’s use AI to make participation in Wikimedia
EASY by reducing friction and helping users onboard
FUN by enabling creativity, exploration and personal expression
MEANINGFUL by becoming deeply personalized and an anchor for belonging to communities that pursue collective purpose 

I look forward to continuing the dialogue started in Frankfurt. Let’s start deliberating and strategizing about the next generation of Wikimedia together. Let’s start by dreaming about a Wikimedia cathedral build on the foundation of the open bazaar.