This is my quick attempt at formulating Francisco’s (@xiq on Twitter) vision behind the Epistemic Garden, the organization behind the Community Archive project.
Large groups of people faces a trade-off between structure and authenticity. Organizations with thousands of people, like large companies, coordinate by defining tasks based on the needs of the structure and allocating people to fit those tasks.
On the other hand, in online communities where every member follows their internal drives faithfully, it is hard to make the contributions of the members compound into something bigger than individual efforts. We even have a saying for that: it’s like trying to herd cats.
Organic contributions point in different directions and cannot efficiently build on each other to achieve deeper outcomes. In large enough groups that gather around shared interests, progress does happens though. New memes get crafted and build upon each other, new useful words get introduced and become widely used. People meet, start projects, and may even found new organizations based on their shared visions.
But these processes happen with little intentionality and leave a lot of value on the table. Not value in the sense of economic value. We don’t want to project the corporate goals of a structured organization like a company onto a group of dreamers/hobbyists/thinkers/builders. Without bending the authentic drives of its members, organically organized communities have the potential to nurture and fulfill these drives more efficiently.
Think of all the connections between soulmates that will never happen because their social graph is not connected enough, important insights that will never propagate further than isolated discussions, or worse, the community that loses its integrity because of outside commercial or political influences.
I believe there is promising work to be done to break this apparent paradox between scalable organizations that require top-down definitions of direction and organic, authentic but inefficient communities.
A vision for this is “building an exoskeleton for online communities.” This would be a layer of service that helps foster what their members do best, much like how a physical exoskeleton amplifies the movements of its user. This can include:
- Organizing events and meetups for in-person connections
- Matchmaking services to create warm connections
- Sense-making services that help the progress made in online discourse compound over time by mapping discussions (e.g. steel-manning different positions in a debate), introducing new memes, or offering historical analysis.
AI can dramatically reduce the cost of these last two types of services, opening up the possibility of bringing to organic online communities coordination qualities that have until now been reserved for either small or large but rigid structures. Importantly, it is possible to create AI-powered community services whose quality increases as AI progresses. I think this is an exciting opportunity to develop thick practices around AI tools, converting raw AI capabilities into community strength.
Before building the services, the first priority is having access to the data produced by the community. This is the goal of the Community Archive, that nows count 7M tweets.
On top of this rich collections, here are two examples of prototypes that prefigure what this exoskeleton could look like. First, Magic Search build by Sofia Vanhanen is a matchmaking app using embedding search based on tweet history from the Community Archive.
Second, Epistemic Garden’s prototypes of trend visualizers and embedding maps aim to provide the sense-making services. A longer-term vision could be to create researcher agents that are able to produce discourse mappings of the quality of Zvi Mowshowitz’s AI news coverage.
I am currently working with @xiq on a prototype exploring this direction. Stay posted for future updates!
Want to hear when I post something new?