A quick prototype for CCA's upcoming AI lecture series.
California College of the Arts is hosting After Intelligence, a lecture series exploring AI through artistic, ethical, and speculative lenses.
I prototyped this microsite using Claude’s frontend skills in a single afternoon session. Dark minimal aesthetic, responsive nav, custom button system, view transitions.
When a promotion review required a portfolio with minimal notice, I didn't just scramble to document my work—I pioneered a methodology for AI-augmented creative development that fundamentally changed how I approach projects.
My promotion review was coming up, and my manager asked a reasonable question: “Can you share a portfolio that demonstrates the scope of work you’ve been doing?”
Simple request. Totally fair. One problem: I didn’t have a portfolio.
Not just “my portfolio needs updating.” I mean I had nothing. No case studies, no project showcases, no structured demonstration of my work. Building a professional portfolio from scratch typically takes weeks of planning, design iteration, content writing, and development. I had days, maybe a week at most before the review.
This is the story of what happened next—but more importantly, it’s about the methodology I discovered in the process. Because this wasn’t just about building something fast. It was about fundamentally rethinking how humans and AI can collaborate on creative work.
Faced with a tight deadline for something important, conventional wisdom offered two bad options:
I rejected both. Not out of stubbornness, but because I sensed there was a third option that nobody was talking about: What if the bottleneck isn’t time, but how we allocate human attention?
Here’s what I realized: Most development time isn’t spent on high-level creative decisions. It’s spent on:
These tasks are necessary, but they’re not where creative value comes from. The real value is in:
Traditional development forces you to do both. AI collaboration lets you focus on the second while delegating the first.
I didn’t start by planning what to build. I started by having Claude interview me like a recruiter.
The prompt: “Before we build anything, interview me to understand my professional narrative and value proposition.”
What followed was a 30-minute conversation that clarified:
The insight: AI isn’t just for implementation—it’s for strategic clarification. By forcing me to articulate my value proposition conversationally, Claude helped me understand what I was really trying to communicate.
What I learned: This interview approach works because AI asks follow-up questions humans often skip. It probes assumptions. It asks “why does that matter?” and “how is that different from what anyone else would do?”
With clarity on the narrative, we moved to prototyping—but not in the traditional way.
Instead of:
We did:
The process:
The insight: When iteration cycles go from days to minutes, you can explore creative directions you’d normally dismiss as “too risky” or “not enough time.”
What I learned: Speed unlocks creativity. When trying something new doesn’t cost you half a day, you’re willing to experiment. I tested design systems I’d never attempted before because the cost of failure was 10 minutes, not 10 hours.
This is where the methodology got interesting. I wasn’t directing AI to build my vision. I was partnering with AI to explore possibilities I hadn’t imagined.
Example: The modal-based project showcases.
My initial idea: “I want interactive project cards that expand to show details.”
Claude’s interpretation: “What about a SPA-like experience with smooth transitions, modal overlays, and no page reloads?”
My response: “I’ve never built that before—can we try it?”
The result: A sophisticated interaction pattern I wouldn’t have attempted on my own because I assumed it would be too complex for the timeline.
The insight: AI partnership isn’t about speed-running your existing ideas. It’s about expanding the possibility space of what you can attempt.
What I learned: The best AI collaboration happens when you’re willing to be surprised. I went in thinking I knew what I wanted. The partnership revealed possibilities I hadn’t considered because I’d mentally pre-rejected them as “too hard for the timeline.”
The final phase was deployment—and this is where the methodology proved its completeness.
Traditional workflow:
AI-augmented workflow:
The insight: AI collaboration can encompass the entire development lifecycle, not just the coding phase.
What I learned: The methodology works when it removes all the tactical bottlenecks, not just some of them. Deployment is often where timelines explode—days lost to configuration issues, documentation gaps, and platform-specific quirks. AI eliminated that entirely.
1. Chunked conversations with clear goals
2. Strategic briefs instead of technical specs
3. Rapid feedback loops
1. Perfect code quality
2. Comprehensive testing
3. Deep customization
This methodology isn’t just about portfolios. It’s about a fundamental shift in how creative professionals can work.
The old model: Human does everything, slower but controlled The new model: Human focuses on strategy/creativity, AI handles implementation, faster and often better
But here’s the nuance: This only works when you’re clear about what AI can’t do.
AI can’t:
AI can:
The key: Keep strategic control, delegate tactical execution.
Start with even more strategic clarity - The interview phase was valuable, but I’d spend more time on competitive analysis and differentiation strategy before building
Build in quality checkpoints - Schedule explicit moments to step back and assess: “Is this still aligned with my strategic goals?”
Document the methodology as I work - I discovered this process in real-time, but having it explicit would make it reproducible
Test assumptions earlier - Some design choices I’d have validated with quick user feedback before committing to implementation
The ironic thing? This blog post exists because the portfolio worked too well as a demonstration of AI collaboration—and that became its weakness as a portfolio piece.
When people saw the portfolio, they thought: “Did Joel really build this, or did AI do everything?” The very thing that made it impressive (rapid AI-augmented development) raised questions about my actual capabilities.
That’s why it’s now a blog post instead.
As a portfolio piece, it was self-referential in a weird way. As thought leadership about AI-augmented workflows, it’s actually interesting. It shows not just what I can build, but how I think about collaboration, methodology, and creative work.
And that’s ultimately more valuable than any portfolio piece could be.
The 24-hour portfolio sprint taught me something crucial: AI doesn’t replace creative work—it amplifies creative capacity by removing tactical bottlenecks.
The portfolio itself? It helped make the case for my promotion. But more importantly, it gave me a methodology I now use for all creative projects:
This isn’t just about building portfolios faster. It’s about building anything faster while maintaining—or even increasing—creative quality.
The question isn’t “Can AI build this?” The question is “Where should I focus my irreplaceable human attention to create the most value?”
Everything else? That’s what AI partnership is for.
Want to explore this methodology further? I’m continuing to document my experiences with AI-augmented creative workflows. Check out my other posts about AI learning reflections and thought leadership on modern development practices.
Have questions or thoughts on AI collaboration? I’d love to hear your experiences. Connect with me through the contact form or find me on the platforms listed there.