#blog #writing #publictext #knowledge-graphs #social-software #commons #info-politics
We drastically under-share knowledge given the technological potential. This isn’t a tools problem. We have more tools for communication and publishing than any generation before us. It’s a design problem. The tools we have are optimised to capture, not to share.
The platforms that dominate our digital social lives are extraordinarily good at extracting signal from our activity. Every click, dwell, connection, and query feeds rich knowledge graphs - models of who we are, what we want, and how we relate to others and to ideas. These graphs capture intent, alignment, connection, context. This is genuine collective intelligence: the patterns that emerge when millions of people interact, curate, and make sense of the world together.
But we don’t get access to those graphs. What we get back are timelines: shallow, ephemeral, algorithmically sorted for engagement rather than understanding. No persistence. No organisation. No way to build on what came before.
The knowledge graphs platforms construct from our activity are genuinely impressive. They encode not just explicit connections - who you follow, what you liked - but implicit ones: what you lingered on, what you skipped, what you searched for after reading something, who you engage with most, what topics cluster around your attention.
These graphs capture intent (what you’re trying to do), alignment (what you value, what you agree with), connection (who and what you relate to), and context (when, where, how you engage). This is rich structured knowledge. It’s the kind of information that makes things findable, relatable, and useful.
And platforms have proven it works. Recommendations, search ranking, ad targeting, content moderation-all of it runs on these graphs. The signals we generate are legible, capturable, and valuable.
The problem isn’t that this is happening. The problem is the asymmetry in who benefits.
The graph is the product. We generate it; they own it; we don’t benefit from it - except indirectly, and mostly in ways we didn’t ask for.
The primary use of all that captured collective intelligence? Ad targeting. Occasionally recommendations that surface something useful, but more often engagement optimisation that serves the platform’s metrics, not our understanding. This is a pathetically low use of what we collectively produce.
Meanwhile the graphs themselves remain proprietary. You can’t export your corner of the graph. You can’t query it. You can’t build on it. You can’t even see it. The knowledge you helped create is inaccessible to you, and the connections between your contributions and everyone else’s are locked away as a competitive asset.
We’ve built the most sophisticated infrastructure for capturing collective intelligence in human history, and we’ve pointed it at selling ads.
The same signals platforms extract implicitly could be laid down explicitly in public infrastructure. Not captured - contributed. Not proprietary - shared.
This isn’t speculative technology. Platforms have already proven these signals are valuable and capturable. The graph structures, the relationships, the contextual metadata-all of it is well understood. What’s missing is the intent to build it as a commons.
Imagine social software designed around organised, open knowledge sharing. Where the knowledge graphs are publicly accessible. Where intent, alignment, connection, and context are laid down deliberately by contributors who understand what they’re building together. Where the accumulated structure is something we can query, build on, and use-not just something that happens to us.
The same activity we currently give away for free to closed platforms could instead feed shared infrastructure that compounds over time. Collation, filtering, curation, connection - all of it contributing to a public resource rather than a private moat.
When knowledge graphs are public and accessible, they become infrastructure for action - not just consumption.
We could use our collective collation and filtering to actually build things: tools, resources, shared references that persist and improve. We could organise ourselves better - finding alignment, coordinating effort, identifying who’s working on what and where the gaps are. We could accelerate progress by building on accumulated knowledge rather than rediscovering it, by making connections visible rather than hoping algorithms surface them.
This is what social software could be. Not timelines optimised for engagement, but structured collaboration optimised for understanding. Not ephemeral feeds, but persistent, organised, queryable knowledge that grows more valuable as more people contribute.
The difference between what we have and what’s possible isn’t technological. It’s about what we’ve decided to build-and who we’ve decided should benefit.
Not technology. The hard problems are solved - platforms have demonstrated that comprehensively. We know how to capture these signals, structure them, make them useful.
What’s missing is intent. The will to build public infrastructure instead of private platforms. The coordination to make it happen. The recognition that collective intelligence deserves better than ad sales.
This isn’t a call for better platforms or nicer corporations. It’s a recognition that the knowledge we generate together could be ours - structured, accessible, and useful. That the graphs we build through our activity could be commons rather than products.
We have the technological potential. We just haven’t decided to use it this way yet.