I asked an AI to help me reimagine my personal site. Not redesign — reimagine. I didn’t describe a vision or list requirements. I just said “reimagine what a personal website can be.”
The AI didn’t generate ideas for me to evaluate. It asked questions. Multiple choice questions, one at a time. And my answers — quick, intuitive, sometimes combining options — accumulated into something I wouldn’t have designed if I’d sat down to design it.
I started with six words: “reimagine what a personal website can be.”
The response wasn’t a mood board or a list of trends. It was a question:
What feels limiting about the current “sections on a page” format?
- It’s static — visitors scroll, consume, leave
- It’s one-dimensional — everyone sees the same thing
- It doesn’t capture who I actually am
- It’s disconnected from my actual work
I picked 3. Not because I’d thought about it before — I hadn’t. But seeing it phrased that way, I recognized it.
Then: When someone spends 5 minutes with you in real life, what do they pick up on that your site misses?
Options included how I think through problems, my taste and opinions, my sense of humor, what I’m currently obsessed with.
I said “1 and 4” — how I think, and what I’m into right now.
A few more questions. Temporal layers. A living workspace. AI-seeded, human-curated. Each answer narrowed the space until a coherent vision emerged that I couldn’t have articulated at the start.
Multiple choice questions work differently than open-ended ones.
When someone asks “what do you want your site to be?” your brain reaches for existing frames — portfolio, blog, landing page. You describe variations of things you’ve seen.
But when you’re handed four options you didn’t write, you’re forced to react. You feel which one resonates before you can explain why. And then you’ve committed to something — not a final answer, but a direction that constrains the next question.
This is how the temporal site emerged. I picked options that felt right. “Archaeological layers” and “AI-seeded curation” emerged from the choices — not from a concept I started with. The coherence came from the accumulation of constrained choices, not from a vision I had at the start.
Design research has a name for this — structured ideation, maybe. But experiencing it with an AI collaborator felt different than a human workshop. No social pressure. No one invested in their own idea. Just questions, answers, and an outcome that surprised me.
Two hours later, I had a design document, an implementation plan, and twelve committed tasks building a site that treats time as the primary axis. Tools I’ve used, projects I’ve worked on, interests I’ve followed — all placed on a timeline you can scrub through like a video. AI extracts signals from my blog posts; I annotate and correct on top.
Is it better than a normal portfolio? I don’t know yet. But it’s mine in a way that feels different. Not because I designed every pixel, but because every choice in the brainstorming session was a genuine reaction. Every choice was mine. The AI just kept asking until something clicked.
That’s the thing worth noting: the creative process with AI isn’t about generation. It’s about structured discovery. The questions shaped the answer — and the answer was something I didn’t know I wanted until I saw the path there.