Field Note: Emergence, Fragmentation, Standardisation
I’ve seen this pattern repeat across different layers of the industry. Not identical each time. But recognisable enough to call. When something new appears, something that changes how systems or users interact, we don’t standardise first. We start.
This tends to unfold in three phases. First, something becomes possible. There is no standard. No agreed interface. Just a new capability. Then, early adopters move. They build whatever works, proprietary data models, custom integrations etc. Nothing aligns. Everything is slightly different. But it proves value. And then, as usage grows, pressure builds. Interoperability matters. Trust matters. Scale matters. Consistency and repeatability matters. That’s where things converge.
What’s often missed is how that convergence actually happens.
There are two common paths. Sometimes, the industry comes together and defines it. A body forms. A spec is agreed. Versions are published. The web stack is full of these moments. Other times, one approach simply wins. Not because it was formally chosen, but because it was used. Widely, quickly, and effectively enough that everything else orbits around it. At a certain point, adoption is standardisation.
The complexity doesn’t sit at the beginning or the end. It sits in the middle. This is the phase where:
- Multiple approaches compete
- Nothing quite fits together
- Teams choose between moving fast or waiting for clarity
It feels messy because it is. But it’s also where the shape of the future gets discovered. This isn’t about saying standards should come earlier. Or that fragmentation is a failure. Fragmentation is often necessary. It’s the exploration phase. The part where the industry learns what actually works.
The question isn’t whether things will standardise. It’s: What pressure will force convergence, and which path will it take? A formal standard? (HTTPS, SMTP...) Or a de facto one emerging from usage? (MCP...)
We’ve seen this before. Different technologies. Same underlying motion. The useful question isn’t what is happening. It’s: Where are we in the pattern, and what does that mean for how we build today?