Whatever it is that does or is going to do AEC thought —people, or AI, or both— maybe we ought to think a little (now seems a good time) about what AEC thought is.
What are we thinking about?
If you like questions like this because you think they might lead somewhere, see here:

As a Bentley BIM user since 1998, I wrote a document in 2007 called “Building Manifest” and sent it to Bentley. That led to employment there, this work getting done, and 8 other software companies developing it since then:
But it didn’t reach Revit, so most people don’t know about it. And it’s not a part of the AEC tech discourse at all.
Here’s the document I wrote in 2007. I found it in a 19 year old email. It’s cringy in a few places. I’d correct some of the phrasing and omit some of the flattery, but it’s as relevant as ever:
What’s been done since proposing this is OK but not nearly enough, and again, hasn’t been seen by the vast majority of users. But here’s a review, and an outlook:
The Tangerine site also includes purpose (like, what are we even doing (or not doing)? And why?:

There is way more to do.
And everyone knows it, really, which you can see in what people actually do, not what they say.
The article “BIMposter Syndrome” by Clifton Harness of TestFit is essential:
https://www.testfit.io/harness-architecture/bimposter-syndrome
This is just cold hard reality that demands an adequate response from AEC software development. And it’s just not there.
I just add that CAD drafting app development hit product-market-fit perfectly (see Clifton’s article) not because CAD companies understood AEC process, but only because they studiously copied the visible expression of it.
BIM developers likewise have had no idea what AEC process is either while, no matter, they evangelized changing the process.
Changing the process to what is addressed only with word salad1, and from what, is never mentioned.
No wonder everything’s stuck.
Here’s the process:
https://tangerinefocus.com/wp-content/uploads/2026/01/look-with-intent-and-show.1.png
- second definition here:
↩︎

