The Difference a Year Makes

Not long ago, Will Smith eating spaghetti was somehow the benchmark for generative AI. A nightmare-inducing video baffled us all and became proof to many that generative AI was all smoke and no fire.

Then it got better. And then better again. And suddenly, this past year spawned entire industries dedicated to answering a single question: what’s real anymore?

The verdict is in. We’re living in a new era of work and automation. And most corporations seem to agree.

But I’m not here to talk about AI in the abstract. I’m here to talk about how it changed my work—and maybe your work too.

Flashback Time

In January 2025, inflation was still top of mind, and wages were still a source of stress for millions. I couldn’t solve that problem, but I knew how to help people understand it.

So I designed an inflation calculator using data from the Bureau of Labor Statistics. You’d enter your starting salary, when you started, and what you make now. One API call later, you could see whether you were actually earning more, or less, once inflation was factored in.

Simple idea. Big impact.

I googled. I Stack Overflowed (RIP soon). I tried my best. I failed every time. Normally, I’d ask an engineer friend for help. I did this time, too. They were busy. All of them.

I felt stuck.

Then someone asked, “Have you tried Cursor?” I hadn’t. In fact… what was Cursor?

When Everything Changed

Fast forward to March. I was in a hotel room on a cold, miserable day in Chicago. No desire to explore the city, plenty of desire to tinker.

So I downloaded Cursor and started explaining what I wanted to build.

I gave it screenshots of my designs, a link to the BLS API docs, and asked for help. With a bit of back-and-forth, it worked. Not only did it work; I finished the entire project and launched a live version in under three hours.

Granted, it was a simple site. But that same site had already taken me months without success. And now it was live before bedtime. I didn’t even have to stay up late.

Looking back, it’s easy to see that as the inflection point. Nothing about my work, or my career, has been the same since.

Curiosity, Unleashed

After that, I couldn’t stop building.

Ideas I’d shelved for years suddenly felt possible. New ones showed up almost weekly. I wanted to see how far I could push what one person could do with the right tools. Keep experimenting.

At work, I shared what I was building. It felt exciting, but also know it was self-contained. These were just personal experiments, after all. I was fascinated by creating with AI, but not yet fully applying it to improve how I worked. That’s when I knew it was time to explore deeper.

This is where the real learning began.

Experimentation is fun. Impact takes intention and strategy. So I started studying how other people, other experts, were applying AI to their workflows.

The Productivity Shift

I began by obsessively checking AI news feeds, online courses, reading substack, all hunting for practical tips. Eventually, I automated that too. Because of course I did.

But more importantly, I tried applying what I found. Most ideas went nowhere. A few stuck. Those became part of my daily workflow and got shared out to the rest of the organization.

It started with automating routine tasks: summarizing meetings, synthesizing research, analyzing survey responses. Helpful, but small and incremental.

The real shift came with code.

When my company floated the idea of trialing Cursor for engineers, I begged for early access as a designer.

The Cursor Figma Loop

Of everything I’ve automated, prototyping has had the biggest impact.

High-fidelity prototypes take time to produce in Figma. Especially when you’re building entire pages just to showcase a single feature. Then you do it again and again to make it clickable. The result is usually… fine. But it’s not a real working website.

Now I only design the piece I need. I design the feature and its components, then build the full site in Cursor. I don’t waste time mocking up artifacts I don’t need. I test against a real, working version of the product. I iterate directly in code. When it’s done, I bring the final implementation back into Figma as the source of truth.

And I hand engineers real code.

This reduced my time-to-test by roughly 80%. What used to take a week now takes a day. The workflow worked so well that we’ve started rolling it out as the default approach for our design team.

One Year Later

That brings me to today.

I’m no longer begging for engineering help. I’m no longer shelving ideas. I’ve shipped entire iOS apps solo; from idea to App Store. I’m still experimenting, still learning, still adapting as the landscape changes. You can never stop learning in the rapid environment of AI.

But what matters more is this: I’m now the one being asked for help.

I host weekly training sessions. I help teammates adopt Cursor. I’m part of conversations about AI wins and the future of work. The experiments I once did at night are now part of my daily job.

A year ago, I was helpless to bring an idea to life.

A year later, I’m teaching others how to do it.