We live in exciting times. Artificial Intelligence has made programming more accessible than ever before. What once required deep focus and late-night debugging sessions can now be generated with a few prompts and some optimism. It’s revolutionary — and somehow, also, a little depressing.
Those of us who started coding because we loved building things are starting to feel tired. Not the “I need a vacation” kind of tired. The existential kind.
A while ago, I read that many experienced developers — especially those working in agencies, consultancies, and digital factories — are living in a state of permanent burnout. It’s not surprising. We stopped being builders and became feature delivery units.
Instead of creating, we deliver. Instead of exploring, we execute. Every decision must fit neatly into a Jira ticket, wrapped in “agile” terminology that somehow manages to remove all agility from the process. We measure velocity, not value. We chase deadlines, not delight.
The Rise of the Architecture Astronauts (and Their AI Descendants)
A long time ago, I read Joel Spolsky’s classic piece: Don’t Let Architecture Astronauts Scare You.
He described a certain kind of engineer who’s so disconnected from actual implementation that they float into orbit — talking about abstract architectures, grand strategies, and futuristic patterns no one can actually ship.
We’ve all met one. Maybe we’ve even been one, briefly. It happens.
But lately, I think we need a new term — because the AI era has birthed a new archetype.
The AI Astronaut.
This is the person who believes that large language models are the ultimate truth, that prompting is the new programming, and that soon all human creativity will be automated away (preferably by a tool they just demoed at the last all-hands).
They don’t write code — they declare intentions. They don’t understand systems — they orchestrate miracles. And if you dare to ask how something actually works, they’ll send you a link to an article.
Sometimes, they’ll even say things like “we’ll have this feature shipped with AI in a couple of weeks” — without realizing that real systems have dependencies, side effects, and constraints that no prompt can handwave away.
And that’s where the danger lies — not in the tools themselves, but in the illusion that complexity can be wished away.
The Craft Is Still Worth Defending
Don’t get me wrong: AI is incredible. It amplifies what we can do, accelerates our workflows, and makes the entry barrier to programming almost nonexistent. That’s progress.
But it’s also creating a generation of developers who no longer want to understand how things work — just how to get things done.
And maybe that’s fine. Maybe software is supposed to become more like plumbing — efficient, invisible, reliable. But for some of us, the joy was never in shipping the feature. It was in understanding why the feature worked.
So yes, celebrate the AI revolution. Use the tools. Automate the boring stuff.
But keep your feet on the ground. Don’t float away with the AI Astronauts.
We already have enough of those in orbit.
Because underneath all the frameworks, prompts, and plugins, there’s still a craft worth protecting — the quiet satisfaction of understanding a system deeply enough to make it dance. AI can assist, but it can’t care. That part is still up to us.
Top comments (0)