I spent twenty years acquiring a skill set that took a lifetime to build and that no one will ever need to compile again.

How I almost missed it

I don’t have the stereotypical programmer’s origin story. No disassembling computers at age seven, no BASIC programs typed into the family TV. My path to programming had several near-misses, and without a slight nudge (internal or external) at any one of them, I might have ended up somewhere else entirely.

Age 6

My father, a technology-curious lawyer, owned one of the first IBM machines (purchased for something north of $10,000, by family lore). He died when I was six, and the machine sat in a corner until it was outmoded. I was always the kid who fixed the electronics and troubleshot the gadgets, but spelunking in that machine never occurred to me.

Age 11

My aunt (his sister, a trained computer scientist who had left a PhD program) advised us to buy a Mac. When I stayed with her one summer, she suggested I play with HyperCard while she was at work. I made an airplane move around the screen. Fun! But I didn’t get a real introduction to programming, and I didn’t make the connection that what I was doing was exactly that.

Age 16

I joined a FIRST Robotics team and was responsible for the wiring from a schematic an electronics mentor had designed. I attended the programming sessions too, though the programming mentor had written the real code himself and didn’t actually teach it. On wiring day, I did it alone. When a pump didn’t work, I traced it to a bug in the code. The programming mentor swore up and down it couldn’t be a code problem. I made him show me anyway, found the issue (a 0 that should have been a 1 – a classic), and it worked. That was a good day.

Age 19 — found it

I went to MIT because I felt most at home among the people I met there, and because as a low-income student I needed a place that would give me a full needs-based ride. My interests were broad: linguistics, cognitive science, philosophy, classical literature. Everyone takes the same core classes freshman year. Second semester, I took computer science, taught in Scheme (a Lisp dialect). Because Lisp was obscure, the students who’d been “coding since birth” had no real head start. I discovered I was good at it. It was addictive. I was helping more experienced classmates before the semester was out. Course set.

Building the impossible resume

Early in college, I remember staring at a resume full of programming languages and wondering how anyone could possibly accumulate that many. At that point I was still internalizing what ls did at the command line, and my mind had genuinely been blown by a WYSIWYG HTML editor (graphical interface generates HTML, you copy it into nano, you post it to MIT’s public html directory, and suddenly you have a website: that easy, but it didn’t yet occur to me I could craft it by hand).

It turns out: if you do this work 40 to 60+ hours a week, across companies and research labs and side projects that each demand something different, you see it all. The tradeoff is just time. There is no secret.

The only other entities that will ever need, or have the time, to develop this kind of full-stack fluency are LLMs.

I worked up and down the stack the whole time: front-end, back-end, mobile, systems, distributed computing, cryptography, kernel programming. I wrote my master’s thesis in C++, contributed to machine translation research, wrote Fortran modules for my HPC PhD work, shipped iOS apps in Objective-C, and competed in hacking challenges against NASA security teams (and won). I taught Python to high school students, college students, and hobbyists across multiple continents. I wrote a book on quantum computing and another on Python for astrophysics.

The tooling evolved too: nano, pico, emacs, vim (six months, never again), Sublime, Eclipse, VS Code. But in the last 5 years, my previous tools and programming languages seem like raw bits compared to assembly. Now I work in Copilot, Windsurf, Cursor, Claude Code, OpenClaw. Programming by typing, by voice, and above all by vibes. The full arc, assembly to agents.

A privilege to bridge the past and the future

When I was nineteen, that list of languages felt insurmountable. It wasn’t. It was just a career. If you stick with the work long enough, and the projects keep asking for new things, you accumulate it.

I’m watching my own kids start to learn to program: and we switch back and forth, turning off the AI assistant to write a Python loop by hand, working through Scratch, getting early exposure I never had. I’m actively trying not to let them miss the window the way I almost did. That said, I’m aware that most of what I learned wasn’t from formal instruction; after the initial spark finally took, it came from specific projects and jobs that required it. From working fulltime on a craft.

Those of us at this career stage in history, having accumulated more than a decade of hands-on programming, with more than a decade left in our careers, have something unusual: we have the expertise of the past, are building the tools of the future, and will have the expertise of the future. We learned the craft when it still had to be learned by hand. We’re now building the tools that make that unnecessary. I’m genuinely excited for what comes next. And I’m mourning the end of a handcrafted art form that won’t come back.

What should we call what we are becoming? Programmers is not the right word. Builders.

A postscript image as an illustration of the new age

I asked ChatGPT to generate an image for this blogpost and it came up with:

Definitely doesn’t quite look like me. When I followed up with a prompt: “I’m a girl” it came up with:

a little garbled but closer to the truth.

But with the right prompt engineering (my trick is to ask the AI to write its prompt for my review, iterate on it several times, before finally turning the prompt into an image), I got something highly professional.

The new tools also require craft, but the craft is collaborative. They require taste, but its taste in cooperation with a different intelligence. And mostly they require the urge to make and the determination to make it better.