Big Tent vs. Little Tent

The idea behind the big tent vs. the little tent centers on how the profession of architecture adapts to changes in the professional landscape. With the recent proliferation of Artificial Intelligence (AI) — starting with language and image models, then moving quickly into other productivity tools — architects need to become aware of how we can leverage AI without falling behind. I don’t want to jump on the bandwagon of doom and gloom with “adapt or die” messaging, but after some personal and professional exploration of AI tools, I can see both future conditions clearly.

To adapt to new technology like AI means to experiment with it and find your own usefulness within it. Much like AutoCAD improved productivity and speed over manual drafting, and BIM further accelerated 2D CAD while adding 3D modeling to our regular workflow, AI is evolving at a pace that can dramatically change how we practice architecture. I have gone from passive dabbler, to curious learner, to tool user, to the interesting creative practice of toolmaking.

Toolmaking 101

I have begun developing my own software to handle very custom tasks that I see a need for in my practice and academic research. Using AI for toolmaking is marrying my curiosity and skillset with the design thinking that was embedded in my architecture education. The iterative planning, process, and production loop of design thinking feeds directly into my toolmaking projects in AI.

The antiquated term Master Builder — what architects used to be known for — has long been relinquished over the past 80–100 years through lawsuits, liability, contractors, accountants, and a host of other professions that found ways to edge in on the role of the Architect. Now marginalized to a necessary evil by many clients (perhaps this experience is solely mine, but I would bet money that if you are an architect reading this, you have felt this pressure too), toolmaking with AI is an opportunity for architects to level the playing field and reestablish our role — not as Master Builder, but as the conductor of the orchestra.

How I Got Started

Like most of you, I first heard about AI — in particular large language models (LLMs) like ChatGPT from OpenAI — and initially thought, “what does that have to do with architecture? Pass!” I tried out a couple of prompts, got some answers, and ultimately put it aside without understanding the value. The mistake I made, as I imagine many of you do as well, is treating LLM prompting like a Google search. That is the wrong mental model.

I also notice my university students making this same mistake. They get called out for using AI (even at a technology-focused university) and become afraid to admit when they’re using it — if they touch it again at all. Working with AI requires a bit of rewiring: shifting from search queries to conversing with a capable assistant, one that can be trained and guided.

Remember: this version of AI is the worst version you will ever use. AI is constantly evolving, and with each new release, models are gaining the capacity to become better assistants and take a more active role as a team member in your practice.

Claude Code

The next big “aha” moment came when I took the leap of faith to experiment with Anthropic’s Claude model — in particular, its computer terminal interface called Claude Code. Yes, I said terminal interface. Get over it. I did. Once you read the instructions on the claude.ai documentation page, getting up and running with Claude Code is far less painful than the words “terminal interface” sound. There are numerous YouTube tutorials that can ease you in, and I’ll share some links below.

Claude Code is really where the magic happens for toolmaking. The biggest difference I’ve experienced compared to the web interfaces of ChatGPT or Claude.ai is the ability to work directly with files on my own computer. Instead of copying and pasting code snippets back and forth, Claude Code operates inside your actual project — reading files, writing changes, and building software alongside you in real time.

The Bigger Tent

So what does any of this have to do with big tents and little tents?

The little tent is traditional architecture practiced by educated professionals content to stay in whatever lane they have carved out for themselves — lanes often determined by other professions carving away at scope. The big tent is about expanding our network beyond our own kind: including allied practitioners like graphic designers, interior designers, and others who already appear on a typical drawing set titleblock. But the real opportunity is larger still — expanding to include software developers, UX/UI designers, fabricators, blockchain practitioners, augmented reality designers, and many more areas previously seen as unrelated to architecture.

AI-empowered toolmaking extends our own training far beyond the part-time coder who used to write small scripts in their spare time. It enables a powerful practitioner who builds custom design tools to provide more impactful service at a fraction of the time and expense. The architect as toolmaker isn’t abandoning the profession — they are expanding it. The tent gets bigger every time we reach across a disciplinary boundary and bring a new capability back into our practice.

The question is not whether to adapt. The question is how wide you are willing to make your tent.