Manifesto

The rapid advancement of AI is driving change in every profession in the knowledge economy. Architecture is not an exception, and we wouldn't want it to be.

We've felt it firsthand. AI changed how software gets built, in our own team, over the last twelve months: mundane tasks that used to take days to weeks reduced to hours, and the value of a software engineer moved from output to judgment. We're not doing anything particularly special. But if we hadn't leaned in, we'd be behind today, and behind in business eventually means out of business. Lawyers are living the same shift right now. So is every profession built around knowledge and intelligence.

Here's what we believe:

The architect becomes more valuable, not less.

AI is getting very good at the procedural work. It can amplify your creativity in ways that will genuinely be useful and fun. Over time it will even get better at judgment, task by task. But the tools don’t hold: the whole project, the relationship with a client, the ability to stand on their site and imagine their life there and the accountability that comes with your name on the drawings. When production gets cheap, your judgment goes where it matters most, and judgment and trust are what clients are really paying for.

Adopt technology for your clients, not for its own sake.

The gap between practices that adopt well and practices that wait is already opening, and it will compound. But adoption only counts for what it delivers: more directions explored while ideas are still cheap to change, decisions made at the speed of conversation, clients participating as actively as they want to, from wherever they are. Fewer changes later. Fewer regrets, in concrete. And something bigger: for most clients, building is a once-in-a-lifetime privilege. It should feel like one. The practices that turn the journey from a source of anxiety into the experience of a lifetime will never be short of work.

Curiosity beats certainty.

If you feel both excited and uneasy about AI, your instincts are working. Nobody knows exactly where this lands, including us. But there is one clearly losing move, and it's waiting to see how it plays out. Or worse, dismissing AI and not good enough to do your job and more specifically, aspects of your job. The practices that thrive won't be the ones that picked the perfect tool. They'll be the ones that built the habit of experimenting, adapting, and staying in motion.

The exploration layer belongs in the browser.

Your CAD is where the building becomes precise: modelled, coordinated, documented, consented. It's the source of truth for the project, and Harth works with it, in and out, nothing lost in between. But around that precision sits a different kind of work: exploring directions, visualizing, deciding together with clients and consultants. That work is collaborative by nature, and collaborative work belongs on the web: multiplayer, link-shareable, connected to the world. Beside your CAD, not instead of it.

Nobody can keep up with AI alone.

New models, new tools, new claims every week, most of them louder than they are useful. We work with this technology every day and even we sprint to stay current; you have a practice to run and clients to serve. So that's the partnership. We live at the frontier and separate what's real from the noise. We turn what's real into workflows that make your practice genuinely more productive, inside your actual projects. Your part is the curiosity and willingness to try new ways of working adapt your workflow. Ours is making it count. Join us on this journey.

This is the biggest wave of change ever to hit knowledge work, and it will not skip architecture. The practices that meet it with curiosity will come out the other side more capable, more valuable to their clients, and more resilient than ever.

Built to adapt. Built to last.

Cheers,

Scott and Tom