There are way more books than editors.
That’s not a complaint—it’s a model I keep coming back to.
AI is starting to feel like the self-publishing revolution for knowledge work. The tools now generate a lot of output. Code, docs, analysis, mockups—you name it.
But if everyone has a book, the value shifts to the person who knows how to edit. As engineers, how can we stay relevant with AI?
The New Shape of Work
I don’t think AI is taking jobs away wholesale. It’s just quietly changing what the job is.
Suddenly, “doing the work” looks more like:
-
prompting a first draft
-
editing what comes back
-
deciding what’s good enough to ship
The bottleneck isn’t output. It’s judgment.
Not All Work Is Equal Anymore
I’ve been thinking of AI-era work as a kind of funnel:
Layer | Role | Future Value |
---|---|---|
Upstream | Framing the problem, defining direction | 🚀 High |
Middle | Generating raw output | 📉 Shrinking |
Downstream | Reviewing, refining, validating | 🚀 High |
In the middle tier, AI’s getting faster. But upstream and downstream still need humans with context and taste.
And that’s where things get interesting… and honestly, a little uncomfortable.
What If Even That Isn’t Enough?
I’ve always leaned toward the upstream side—mapping patterns, breaking down problems, making architecture legible. It’s work I like. But lately I’ve been asking:
In a world where more people are doing this kind of work… is it still enough to stand out?
There are a lot of sharp engineers who can define systems and edit AI output. And as the middle collapses, more of them will be aiming for the same higher-value work.
It’s not just about being good anymore.
It’s about staying relevant in a world where the definition of “good” is shifting.
What Actually Sets People Apart?
What I’m starting to see is that it’s not the code or even the architecture—it’s:
-
The ability to navigate ambiguity
-
Taste, not just correctness
-
Thoughtfulness about how AI fits into human workflows
-
Influence—helping others make better decisions, not just better code
Those are the things I’m trying to sharpen, and frankly, trying to name. Because the more output we automate, the more we need people who can say, “Here’s what matters.”
This Feels Different Than Other Tech Work Disrupters
I’m not a season veteran software engineer yet – I’ve worked less than 10 years. So, sometimes I wonder—is this how engineers felt when containerization came out?
Probably not. That was a shift in tooling, in how we built and deployed software. It was frustrating at times, but it didn’t question the core of what we did. If anything, it made good engineers more powerful.
This feels bigger.
It’s not just changing how we work—it’s changing what counts as work, and who gets to do it.
There’s something heavier about that.
Less technical, more existential.
What I’m Trying to Do Differently
This is where I’ve landed, for now:
-
Keep focusing on judgment-heavy work, not throughput
-
Write more—because writing forces clarity and scales influence
-
Be thoughtful about how I use AI, not just whether I use it
-
Look for problems that don’t have clean owners—especially across team boundaries
-
Stay honest about where I’m coasting and where I need to grow
Still Figuring It Out
I don’t think I have the answers yet. But the question that keeps me moving is:
If AI can do most of what I used to do—what’s left that only I can bring?
That’s the edge I want to build on.