AI Makes AEC Data Computable

After NXT BLD and NIBS, I left with the feeling that something has shifted in AEC.

Every firm seems to be talking about data now. Some are building internal tools. Some aretesting AI workflows. Some are trying to connect drawings, models, documents, schedules, and project knowledge in ways that were difficult even a year or two ago.

But I do not think the interesting question is only “how do we use AI?”

The deeper question is: how much of AEC work can finally become computable?

For a long time, we have had digital files without truly computable information. A PDF isdigital. A drawing set is digital. A BIM model is digital. A spreadsheet is digital. But thatdoes not mean the knowledge inside them is easy to search, test, compare, or reuse.

This is where AI changes the situation.

On one side, AI can help make previously hard-to-compute AEC information more accessible. It can extract information from documents, structure messy inputs, summarize project history, compare requirements, and help turn scattered knowledge into something we can work with.

On the other side, AI lowers the barrier to computation itself.

Before, if an architect, engineer, or project manager had an idea for a tool, they often needed to wait for a developer, a software vendor, or a feature request. Now, more people can begin turning their own questions into small scripts, checks, workflows, and tools.

That matters.

It is strange, when you think about it, that so much of our work is trapped inside softwareinterfaces. We click buttons, move through menus, export files, and wait for platforms todecide what is possible.

But the actual value is not the interface.

The value is in the ideas, rules, geometry, constraints, relationships, and decisions behindthe work.

If a room has requirements, we should be able to test them directly. If a specification defines constraints, we should be able to compare against them. If a coordination issue keeps appearing across projects, we should be able to detect the pattern earlier.

This is why we are hosting a free one-hour workshop next week:

[Data Science Fundamentals for AEC - Build Your own tool with AI · Luma Most AEC professionals know AI matters, but it is not always clear what is actually useful in day-to- day work. In this free 60-minute online session, we will… luma.com/datadrivenaec

](https://luma.com/datadrivenaec)

Data Science Fundamentals for AEC - Build Your Own Tool with AI Thursday, June 115:00-6:00 PM GMT+2Online via Google Meet

We will cover the data science fundamentals that matter for AEC, open a terminal together, and build a simple working tool with AI.