I wrote recently about where AI belongs in our design process and where it very much doesn't. This one's about the business side, because lately the questions have gotten pointier.
The one I hear most: since you're using AI, surely your costs are coming down?
Big fat no. If anything, there's a decent argument they should go up.
The work didn't shrink
We're likely spending more time on each project now, not less. What's changed is that the work comes out better and moves faster.
But the actual list hasn't shortened by a single line. Concepts, design, space planning, drawings, drafting, code review, jurisdictional review, pricing review, more drafting, engineering coordination, engineering review, renderings, finish selections, specifications. It goes on. And on.
A different tool doesn't delete any of that. It just means we get through it a little quicker and a little sharper, so projects keep their momentum. Where we might once have told you three months, partly because we're juggling five other things, we can now often say two.
A lot of that comes down to letting AI work while we don't. We'll set renderings running overnight, let things churn while the office is dark, and by morning it's waiting for us. Less about AI being brilliant, more about being clever with the hours.
Please, don't buy every app
Here's my genuine warning, from someone who has not taken her own advice: don't throw money at every shiny new AI tool that appears. I have been an absolute sucker for this. I've tried an embarrassing number of them. A few were terrific.
But the real value was never in the app. It's in working out how this fits your business, your team, your process, your brain. Everyone needs a different kind of help. Not everyone needs code analysis, or drawing review, or rendering support. There's no one-size-fits-all here, and anyone promising you otherwise is selling something.
So we're building our own
We're currently building a team app for internal communication, drawing management, and a client portal. There's no shortage of software out there, but none of it did what we needed for the way our team actually works. So we figured, why not?
Will our software work for everyone else? Almost certainly not. And honestly, that's the whole point.
Where AI has truly earned its seat
The business side is where it shines for us. Running the company, wrangling finances, building internal tools so the team can coordinate properly and actually bring clients and contractors into the fold. As a kind of tireless assistant, it's a joy.
And let's be honest with each other: almost none of us got into design for the math and the spreadsheets. AI's willingness to take that off our plates is unmatched. It sorts purchasing and procurement. It helps us build proposals that are genuinely beautiful. It swallows the tedious things that used to eat an entire evening, and it never once complains.
Where all this is heading
I'm not going to pretend nothing is changing. Parts of our industry will disappear. That's real, and I think the people digging their heels in are going to find it very hard to move their businesses forward.
So we're pivoting. We're building our own processes and tools so we can keep doing our best work and take on more of it, with AI humming away in the background where it belongs.
But here's what I keep landing on. The more of this becomes automated, the more valuable the human part gets. Taste. Judgment. Knowing when to break a rule. Sitting across from someone and understanding what they actually want, which is rarely the thing they said first. I don't think AI is coming for that. If anything, I think it's about to make it the most valuable thing we sell.
The question was never whether to use AI. It's how to use it well, and what you refuse to hand over.
And yes, same as last time: I wrote every word of this myself, then had AI tidy up my typos. Only fair to say so.