Stop Asking AI to Design Buildings

Blog
We are drowning in noise. Instead of asking AI to generate "magic," we should be using it to unlock our firm's collective intelligence. It’s time to move from replacement to partnership and start making Design-Driven Data Decisions.
CBCristóbal Ignacio Burgos Sanhueza and some AINovember 9, 2025
708 words / 4 minutes read

Any sufficiently advanced technology is indistinguishable from magic.

Arthur C. Clarke

I’ve been thinking about this quote a lot lately.

If you look at the current conversation around AI in architecture, it feels exactly like that: magic. It is loud, chaotic, and frankly, a bit overwhelming. You have grand promises on one side and quiet anxieties on the other. It is easy to get caught up in the spectacle of it all.

But we need to ground this.

As much as the tech feels like magic, we have to remember a simple truth: we work for people, not for wizards.

My goal—and honestly, my obsession—is to cut through that noise and see AI for what it actually is: a tool. A profoundly powerful one, sure, but a tool nonetheless.

The Real Assets: People and Projects

So, how do we stop treating this like sci-fi and start treating it like practice? We have to look inward.

Before we worry about the future of tech, look at the foundation of our profession. An architectural practice’s biggest assets have never been their software licenses; they have always been their people and their projects. The real question isn't "What can AI design for me?" (that’s the magic talking). The real question is: "How can AI connect these two pillars?"

For any studio with years of history, the greatest opportunity isn't replacing designers; it is tapping into its own collective intelligence.

Think about it. How do we make sure that a designer who joins the team today can design with the full weight and wisdom of the firm's history behind them from day one?

Protecting the Fragile Idea

This is where the shift happens. I’m not interested in asking an AI to "design a building." I’m focused on building systems that empower our teams to make Design-Driven Data Decisions.

Why does this matter? Because the early stage of design is incredibly delicate. I've always resonated with how Jony Ive describes this tension:

I obsess trying to deeply understand the general nature of ideas... Ideas by definition, are always fragile. If they were resolved, they wouldn't be ideas. They would be products that were ready to ship.

Jony Ive

This is exactly it. We need systems not to replace the idea, but to support it while it is fragile. We need to empower our designers with curated insights drawn directly from our own past work so those ideas have a solid foundation to stand on.

Architecture has never been just about drawing. It is about the tradeoffs and synthesis of knowledge. A drawing is just the final artifact of a thousand invisible decisions. The problem is that in today's world, the knowledge needed to make those decisions is fragmented. It’s buried in scattered servers, old reports, and locked in the minds of senior partners.

This is the problem I want to solve.

Imagine being able to ask your own project archive: "What did we learn about waterproofing details from that project we completed in a rainy climate five years ago?"

This isn't about replacing judgment. It is about informing it.

The Future is Partnership Ultimately, I’ve come to a simple conclusion. AI, by itself, will not create better designers. But it will absolutely empower great designers to create even greater, more human-centric work.

My job is to create a better designer experience. By giving teams the tools to access their own knowledge, we clear the path for them to do what they do best.

It brings me back to John Culkin’s famous line:

We shape our tools and then our tools shape us.

John Culkin

The future isn't a battle against technology. It is a partnership with it.

But here is the catch: we have to be the ones leading that partnership. For too long, we’ve outsourced our process to software giants that don't understand the messy, beautiful reality of what we do. If we want to truly own our work, we have to own the tools that make it. We need to build our own workflows, craft our own systems, and take full responsibility for the digital environments we inhabit.

If we shape these tools correctly now—grounding them in our own data, our own values, and our own way of doing things—they will shape us into better, more informed, and more capable architects tomorrow.

That is the real signal in all this noise.

This post was written with the help of some type of AI.
You probably already knew that, but I thought it was worth mentioning.