The old stuff still works
There's a version of this post where I play it safe. Where I write about the future of design in vague, aspirational terms, name-drop a few tools, and call it thought leadership. This isn't that post.
This is the one where I tell you that I built ericjordan.design using Claude, Vercel, Supabase, GitHub, and ElevenLabs — that it started as an experiment on a random Tuesday, and that it turned into one of the most energizing creative experiences I've had in years. The kind that makes you remember why you got into this work in the first place.
Visual hierarchy. Typographic rhythm. The discipline of knowing what to leave out. No model, no automation, no stack decision changes that. The change comes from the velocity at which you can apply ideas.
“The fundamentals didn't disappear when AI entered the room. They became the foundation for something faster, sharper, and more alive than I could've shipped alone.”
What happens when you stop waiting for permission
I didn't have a roadmap for this site. I had a prompt, a vague idea about chatbots, and genuine curiosity about what was possible. So I started.
Claude helped me think through structure and logic in real time. Vercel shrunk the gap between idea and live chatbot from days to minutes. Supabase gave me a backend without the ceremony of building one from scratch. GitHub kept everything honest. And ElevenLabs — this is where things got interesting.
I wanted the blog to breathe. Not just read, but actually be heard. So I built an audio layer: every post, read aloud with enough texture and warmth that it feels like something, not an afterthought. One idea led to the next, the way good ideas do when you're actually in motion. That's the real unlock, not replacing craft with code, but using both as raw material. When you understand what makes a layout breathe, an AI assistant becomes genuinely useful. When you don't, it just generates noise faster.
“I wanted the blog to breathe. Not just read, but actually be heard.”
The workflow is all of it at once
Here's the honest description: refreshing, difficult, easy, and fun, all simultaneously. That tension is exactly what makes it worth it.
It's refreshing because the creative loop is tighter than anything I've worked with before. It's difficult because taste is still entirely on you, the tools don't know when something is off, only you do. It's easy in ways that feel almost unfair, like how quickly you can go from a rough idea to something deployable. And it's fun because there's genuine discovery in it.
You find things you weren't looking for. A chatbot becomes a brand voice. An audio player becomes a signature. A portfolio becomes a proof of concept. This new AI-orchestrated workflow isn't one thing, it's all of them, running in parallel, pushing each other forward.
The lesson isn't about the tools
It's about what happens when you're courageous enough to start. When you lean into both the old and the new — the rules you learned a decade ago and the capabilities that didn't exist last year — you open space for something neither could produce alone.
Merging foundational design principles with the power of AI, automation, and future-facing processes opens an unprecedented opportunity to build, ship, and deploy tastefully, with intention and creativity at scale. But only if you're willing to try.
Beautiful discoveries live just past the part where you almost talked yourself out of it.
Build something. See what you find.