Rereading the thing you wrote three years ago

Next week marks three years since we published the Code of 24, the twelve principles and twelve disciplines that run everything we build at HAUS XXIV. I wrote most of it in a six-week stretch in the late summer of 2022, working out what I actually believed after years of watching good creatives get worn down by bad systems. I remember finishing it and feeling like I’d said everything that needed saying. Three years, dozens of clients, and a few genuinely hard lessons later, I sat down and reread it front to back for the first time in longer than I’d like to admit.

Here’s the honest version of what I found. Most of it held. Some of it needed the real world to sharpen it before I fully understood what I meant by it. And one whole category of what we do now didn’t exist as a concept when I wrote the first draft. This isn’t a rewrite. The Code of 24 stays exactly as published, because a haus that rewrites its own foundation every time it learns something new never actually stood on one. This is the companion piece. What we’d add, if we were building it today, without touching what’s already there.

The principle that proved truer than I expected

I wrote "risk is the rent we pay" believing it, but I didn’t yet know how often we’d have to collect on it. The version of that principle that lives in my head now comes from a specific Tuesday, sitting across from a founder who wanted the safe logo, the safe palette, the safe tagline, because safe felt like less to defend in a board meeting. We walked her through what safe actually costs. Not in dollars first, in attention. Nobody remembers safe. She went with the version that made three people in the room visibly uncomfortable before it made them proud, and eighteen months later that mark is the thing her competitors are quietly trying to copy.

That’s not a one-off. It’s a pattern I’ve watched repeat across nearly every brand that trusted the discomfort instead of managing around it. The principle wasn’t wrong in 2022. I just didn’t yet know how literally to take it. Risk isn’t a cost you absorb once at the start of a project. It’s rent, paid on a schedule, for as long as you want the work to matter.

“I didn’t yet know how literally to take it. Risk isn’t a cost you absorb once. It’s rent, paid on a schedule, for as long as you want the work to matter.”

David Keyes, Founder & CEO, HAUS XXIV

The principle the real world had to sharpen

"We don’t dictate, we educate" sounded complete to me when I wrote it. It wasn’t. What I hadn’t accounted for is how often education gets mistaken for permission, how a client hears "we’re not going to dictate to you" and assumes that means every idea in the room carries equal weight. It doesn’t. We educate hard, and then we hold the line on what the work actually needs, because a haus that explains itself into a worse outcome isn’t practicing conviction. It’s practicing avoidance with better manners.

It took a handful of projects where we explained too gently and shipped something we weren’t proud of before I understood the gap. Now the principle runs with an addendum I say out loud in every kickoff. We will always tell you why. We will not always agree with you, and telling you why doesn’t mean we’re asking permission to be right. That distinction wasn’t in the original document. It should have been. Three years of practice is what put it there.

What didn’t exist when I wrote this

Here’s the plainest fact about the Code of 24. I published it on October 15, 2022. ChatGPT launched publicly six weeks later, on November 30. I wrote an entire operating system for how a creative haus should think, work, and hold itself accountable, and I did it in the last quiet season before generative AI became something every client, every competitor, and every one of our own people had an opinion about. That’s not a gap I could have closed by writing faster. The thing hadn’t happened yet.

What we built in the years since is the Picasso Standard, our internal rule for how AI sits at this table. It doesn’t run the haus. It doesn’t get credit for craft it didn’t do the human part of. The standard is simple to say and harder to live: the issue was never the tool, it’s misrepresenting how the work got made. We use AI to do more with the people we have, never to do less with fewer of them. If I were drafting the Code of 24 today, this wouldn’t be a footnote bolted onto an existing principle. It would earn its own place among the twelve, because it’s become one of the realest tests of whether our conviction is actually conviction or just a nice thing we say in pitches.

“The issue was never the tool. It’s misrepresenting how the work got made. Use it, respect the craft, be honest about both.”

David Keyes, Founder & CEO, HAUS XXIV

We’re not the only ones who add without erasing

I’m not precious about the idea that a set of operating principles should stay frozen the moment you publish them. Amazon launched its leadership principles with fourteen. The list now runs to sixteen, with additions like "Strive to Be Earth’s Best Employer" and "Success and Scale Bring Broad Responsibility" added as the company grew into problems it didn’t have on day one. They didn’t delete the original fourteen to make room. They built onto the foundation because the foundation was sound enough to build onto. That’s the model. Ray Dalio ran Bridgewater Associates for decades on a codified set of principles he kept testing against reality, not because the principles were weak, but because reality kept teaching him things he hadn’t accounted for yet.

That’s what three years taught me about writing something down and calling it a system. The document doesn’t get to stop learning just because it’s published. We do the learning. The document, when we’ve done it right, is just where we keep proof of the work.

What actually happens next

The Code of 24 isn’t changing. Go read it, it’s still the first thing we ever published on this Journal, and it still says what I meant it to say. What changes is us, every year we hold ourselves against it and find the places where three more years of client work sharpened something we thought we already understood. That’s the actual discipline. Not writing the perfect document once. Being willing to say, three years in, here’s what I’d add, and meaning it enough to actually build toward adding it.

Safe doesn’t survive here. Neither does a set of principles nobody’s willing to keep testing.