You’ve probably seen the viral 3D map of Los Angeles. You know the one: a sweeping, time-lapse animation where 1,129,558 buildings rise from the earth, turning empty fields into a sprawling metropolis by 2026.
It’s mesmerizing. Watching the city breathe and expand triggers a visceral sense of awe.
But look closely at those empty early years. That barren 1880s landscape isn’t a snapshot of an untouched city. The empty landscape isn’t an untouched city; it’s a graveyard of demolished history.
The map only extrudes buildings that are still standing. Every missing structure is a demolition the interface cannot show. You aren’t watching a city get built; you’re watching a city survive.
We are being seduced by vibe-coded visualizations. As one commenter bluntly pointed out, these are the hello world of LLM projects. It’s easy to prompt an AI to draw a map. The actual analytical heavy lifting wasn’t prompting—it was the grueling work of joining county lidar footprints to assessor rolls.
The interface promises objective growth backed by precise public data. But the dataset’s inclusion criteria are an implicit demolition record. The exactness of the rendering hides the fact that absence is doing the analytical work.
As AI makes generating beautiful charts effortless, the actual analytical edge shifts from what we can render to what we choose to omit.
Urban enthusiasts and data practitioners alike need to ask what any visualization chooses to leave out. When you only map the survivors, you erase the neighborhoods that were bulldozed, the communities displaced, and the architectural history paved over.
When a dataset only includes the survivors, it doesn’t show you how a city was built—it shows you what the city decided to keep.
This isn’t just a map of LA. It’s an argument about permanence, renewal, and how official records encode a city’s memory. Next time you watch a city bloom on your screen, remember the ghosts of the buildings torn down to make it happen.
FAQ
Q: What is the key takeaway?
A: See the article.