I read Bill Dollins’ recent post, Post-GIS Revisited, and found myself nodding along — not because it neatly resolves the “post-GIS” question, but because it refuses to.
That alone makes it worth reading.
I’ve known Bill long enough to recognize when he’s not trying to coin a term, but instead trying to surface a discomfort. This post sits squarely in that space.
“Post-GIS” has always been a trap term
The phrase “post-GIS” tends to trigger the same reactions every time:
- panic that GIS is being declared dead
- eye-rolling that this is just rebranding
- or a long thread explaining why real GIS is still important
All of which miss the point.
What Bill is really circling — and I think this comes through clearly — isn’t a world after GIS, but a world where GIS no longer insists on being the front door.
And honestly, that world has been forming for a while.
GIS didn’t go away — it spread out
If you zoom out far enough, the pattern is familiar:
- GIS started as a destination
- then became a platform
- then a service
- and now it’s increasingly just… there
Spatial logic shows up inside data pipelines, APIs, validation steps, pricing models, routing engines, AI training data — often without anyone calling it “GIS” anymore.
That’s not loss. That’s diffusion.
It’s the same arc we’ve seen with databases, networking, even operating systems. Once something works well enough, it stops demanding attention.
The real change isn’t technical — it’s behavioral
One thing Bill’s post nudged me to think about (again) is how much classic GIS tooling still assumes a certain organizational shape:
A knowledgeable human will open the system, notice something spatially weird, and fix it.
That assumption worked when:
- data volumes were smaller
- workflows were slower
- and GIS teams were explicitly in the loop
It works a lot less well when spatial logic is embedded inside automated workflows, cross-functional systems, and near-real-time decisions.
This is where “post-GIS” starts to feel less like a technology shift and more like a responsibility shift.
From expertise as heroics to expertise as design
In older GIS models:
- expertise lived in people
- correctness emerged at the end
- context was tribal knowledge
In newer spatial systems:
- expertise gets encoded upstream
- correctness becomes contractual
- context lives in metadata and workflows
That’s not a downgrade of GIS thinking — it’s the maturation of it.
It echoes ideas I’ve written about elsewhere around metadata as operational infrastructure and why humans are not a scalable integration pattern.
The spatial experts don’t disappear — they move earlier in the process. They design the system so fewer things need rescuing later.
If this feels boring, that’s a good sign
If “post-GIS” feels unsatisfying as a label, that might actually be the tell.
The end state of good infrastructure is boring:
- no heroics
- no emergency map fixes
- no late-night Slack threads asking why the geometry looks wrong
Just systems quietly doing spatial reasoning as part of their normal operation.
Bill’s post doesn’t declare an ending. It sketches a trajectory. And I appreciate that restraint.
Because the most honest answer to “Are we post-GIS?” is probably:
We’re post pretending that GIS is the whole story.
And that’s not a bad place to be.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.