Down the rabbit hole.

A year ago, I'd been out of the IDE for over fifteen years. I started my career writing code — mainframe at first, then web through the dotcom — then followed my curiosity into sales engineering, product management, product marketing, anywhere the problem was interesting. The generalist's path. The kind where someone eventually pulls you aside and asks if maybe you should have picked one thing. I'd quietly wondered the same.

Then I ran into Claude at 11pm on a Tuesday and stopped wondering.

I should say up front: I'd never left code entirely. HTML, CSS, JavaScript — still in there. I'd been using ChatGPT for what most people use it for: enhanced Google search, drafting emails faster, debugging things around the house. Quietly useful. None of it felt like a discontinuity. It felt like a slightly better tool.

That changed on a Tuesday.

The night it broke open

I'd been carrying an iPhone app idea around in my head for about three years. The kind of thing I'd sketched in notebooks, thought about on runs, never built. On a hunch, around 11pm, I sat down at the kitchen table and described it to Claude Design. Not a prompt. A description — the way I'd describe it to a designer over coffee.

Minutes later I had high-fidelity HTML mockups I could click through on my phone. Not wireframes. Real screens. I iterated on the brand. I iterated on the features. By 2am the thing I'd been carrying around for three years in my head was sitting in my hand.

I sat back and said "oh — the glass slipper fits."

It wasn't "AI is cool." It was the realization that every "you should specialize" I'd ever brushed off had been wrong. The thing I'd quietly apologized for — being good at too many things — was suddenly the most valuable thing about me.

The next morning I cut over to Claude Code. Phoned a friend (my brother). By the end of the day the app was running on my phone in Expo Go. I haven't stopped pulling that thread since.

The rabbit hole

Since that Tuesday I haven't slept like a normal person. Vampire hours — coding until 4am, sleeping a few hours, getting up like nothing happened. Dozens of commits a night. My wife has been amazing through all of this. She comes downstairs at midnight, doesn't roll her eyes at the demo I'm showing her, says "this is going to be a real business" — and means it. She hasn't seen me this excited about tech in years.

Last week I shipped a feature at 2am that two years ago I would have written a Jira ticket for — directly into the GitHub repo I set up and manage, a repo I'd learned to configure with AI in about an hour (more on accelerated learning in a future post). The velocity is real. Every layer that used to require a specialist becomes something I can move on in days instead of weeks.

But speed without structure is how you paint yourself into corners. That's why my brother — an Enterprise Architect — is in the room. He's not slowing me down. He's making sure the decisions I'm making at 2am don't become technical debt at 2pm. The generalist who can move fast, paired with the architect who thinks three moves ahead — that's the shape that actually works.

The thing nobody tells you

Here is the part I want recruiters and CEOs to actually hear.

I'm not pivoting to engineer. I'm not "vibe coding." I'm a person who started as an engineer, then spent the bulk of my career across sales engineering, product management, and product marketing, and now — thanks to AI — gets to be all of them at once. Those four things in one person, on AI rails, is a fundamentally different shape of operator than any of them alone.

"Jack of all trades, master of none" — the compliment with a knife in it. For most of my career, being a generalist was a soft liability. Recruiters slid me past roles because I didn't slot cleanly into a single JD. Now I can spin up a landing page in an afternoon, code the backend that powers it the next morning, write the launch sequence over lunch, and be selling it to a customer by Friday. Marketing-Chris and engineer-Chris and salesperson-Chris are all the same Chris, working off the same context, with the same taste, in the same week.

That used to take four hires and a quarter. It now takes me and a weekend.

Where this goes

I think we're at the start of a five-year window where the most valuable people in any company are the operators who learned to use this stuff before anyone else figured out it was important. Not the people who built the models. The people who internalized the workflow.

SLED is my proof. Eight weeks, idea to App Store, two brothers. The full story—and how we actually did it—ships when the app does. But this isn't the end of the rabbit hole. It's just the first thing I built after I noticed I was in one.

If you're curious how this actually works—or if you're in the rabbit hole too—say hi. There are more of us than the LinkedIn discourse would suggest, and we should know each other.

— Chris