Software teams just discovered the alignment problem.
Construction has had it for decades.
Maggie Appleton from GitHub Next gave a talk this week that started with a simple observation: as AI makes writing code nearly instantaneous, the bottleneck shifts. It’s no longer “how do we build this?” It’s “does anyone agree on what to build?”
She calls it the alignment bottleneck. When implementation gets cheap, coordination debt becomes the real cost. Teams ships features no one needed. Planning windows collapse. Context lives in people’s heads, not shared systems.
If you work in construction, you recognise that immediately.
Construction’s alignment problem isn’t new. It’s structural. The delivery process was designed around physical and time distance — different firms, different phases, different information environments. Coordination overhead was always the cost of doing business.
What AI is doing now is making that overhead visible to industries that didn’t have to see it before.
Appleton’s prototype, ACE, is a multiplayer development environment where teams and agents work in the same session — shared context, visible prompting history, and collaborative planning before a single line gets written. Not just faster tools. A different surface for alignment.
Now look at what Palantir is doing in AEC.
They’re not competing with Revit. They’re building an Ontology — a persistent semantic layer above design tools that enables AI-driven decisions across procurement, labour, risk, and portfolio data. Once that layer accumulates, switching costs become prohibitive. It becomes how the organisation thinks.
One is a prototype that helps dev teams discover the alignment problem in real time. The other is a company that has spent decades building semantic layers on top of the most complex organisations on the planet — defence, intelligence, and finance — and is now applying that capability to construction.
The construction industry has lived with the alignment problem so long it stopped seeing it as a problem.
The solution may not come from within.