AI Already Took My Job. I Gave It Away on Purpose.

I’ve had a queue in my head my whole life. Problems I can’t yet solve sit there, simmering, while I move on to the next one. Medicine. Art. Code. Design. The medium always changed. The queue never did.

I have a confession. I’ve always been asked to write specs — and I dreaded it. Not because I couldn’t do it. I was good at it. But sitting down for days of heavy, solitary thinking to document what we were building? It was a drain. Every time.

Early on in the AI days, I could see that written specs were a bottleneck, not just for me, but for the entire product and engineering cycle. At the rate engineering was moving, there was no way we would scale or keep up if something didn’t change. That was the real problem to solve.

So I did something about it. I spent months training a system to do it instead. And here’s the thing nobody tells you about that process — I loved it. It was a puzzle. It fed the part of my brain that’s always looking for a better way. The training was engaging in a way the spec writing never was.

That’s when I realized something had fundamentally shifted. Not just in my tools. In my job itself.

Today we sit down as a team, we talk through a feature the way we always have — messy, alive, full of tangents — and we record it. An agent I’ve spent months training takes that conversation and produces a spec we can actually build from. Ninety percent there. The other ten percent is the conversation we needed to have anyway.

That agent didn’t come from a template. It came from months of iteration, of teaching it how I think, what I care about, what good looks like. It’s not replacing my judgment. It’s running on it.

The same is true for how I research now. I used to hire for the parts of research I couldn’t scale — the notetaking, the video rewatching, the pattern finding across dozens of calls. Good work. Slow work. Work that lived or died on whoever I trusted to do it.

Now I come to a customer call with a script in hand and a system that helps me find the patterns across every conversation, every enablement session, every usability study that follows. The calls are still mine. The thinking is still mine. I’ve just stopped losing it somewhere between the conversation and the decision.

I want to be honest about what this means. The notetaking, the pattern finding, the research synthesis — I’ve encoded a lot of that into systems now. I know what that sounds like.

But here’s the reality most design leaders aren’t saying out loud: my team isn’t the size it used to be. Budgets are tighter. Headcount that used to exist doesn’t anymore. The choice was never AI versus researchers. It was: do this work at the quality it deserves, or don’t do it at all.

AI gave me a third option.

Nobody named this quality in me as a strength. For a long time I named it as a shortcoming — it’s literally in my “how to work with me” file.

“My brain never stops. I start things and move on to the next and come back to the first. I call it procrasti-working. When I get stuck I move on, let my subconscious percolate, and come back when there’s a different path. It can look disorganized. It can look like I’m not finishing things. There is a method to it — but I understand if you need to ask.”

Medicine taught me to hold complexity without panicking. Art taught me to sit with ambiguity until something became clear. Design taught me that constraints are where the interesting thinking happens. Code taught me that if you can describe something precisely enough, you can build it.

It took me a long time to stop defending this about myself and start building with it.

The job I used to have still exists. I just don’t do most of it anymore. I built systems to do it better than I could alone, and I moved on to harder problems. That’s not a loss. That’s the point.

What I do now didn’t have a name two years ago. The job isn’t making things anymore — it’s deciding what’s worth making, building the systems to make it well, and staying close enough to customers to know the difference. The work can finally keep up with the thinking.

The same way I always have. Just faster.

I’ve spent my whole career feeling slightly ahead of what the available tools could do. For the first time, that gap is closing.

This is not a cautionary tale about AI and design. This is what it feels like when your whole career turns out to have been preparation for exactly the moment you’re in.

I’m not scared. I’m ready.