It is fascinating to see Andreessen Horowitz's take on one of the biggest problems in the AEC industry, which us design tech nerds have been discussing for decades but never got attention from these major investors. I have so much to say about this article, but here are my 2 immediate reactions.
What I like the most in the article:
Deep diving on MEP– This is spot on. Compared to architecture, MEP is where AI can shine more. There is less subjectivity and taste involved in the work, and it is more mathematics-driven than aesthetics-driven. Also, since the work builds on top of architectural work already completed, it has much clearer context and constraints, as well as a narrower focus.
What this article is missing:
It does not talk enough about why disrupting Revit has been so incredibly difficult. "Autodesk monopoly" is not a sufficient reason to explain this. Deeply understanding this point is critical to any effort to "attack Revit directly." Here is what I think is at the core of it: while Revit sucks in so many ways, it is actually quite good at one thing; the production of drawings. Architects use other 3D tools like Rhino or SketchUp, but these are primarily modeling tools that lack robust documentation capabilities. Architects' and engineers' end deliverable is ultimately 2D drawings, so if you want to make them absorb the switching cost, your product needs to be exceptionally good at drawing production (preferably with the help of AI). Ironically, this is also why AutoCAD remains dominant in the AEC industry. (I actually wrote a blog post about it a while ago. Link in comments if you're interested!)
The only thing that can change this whole narrative is the emergence of robotics in construction. If robots begin to build, perhaps architects' and engineers' means to communicate with builders will change to something else. But that's a story for another time!