I've been writing about how AI changed my workflow, how we're not really replacing ourselves, and how I built tools to bridge design and code. But all of that raises a question I keep coming back to:
What does evolution actually look like for someone like me?
The old ladder doesn't exist anymore
A few years ago, the path was clear. Junior dev, mid dev, senior dev. Maybe tech lead. Maybe architect. Each step meant writing harder code, leading bigger teams, owning more complex systems.
That ladder still exists on paper. But the rungs feel different now.
I don't think my next step is becoming a better coder. I think it's becoming a better planner. Something between a PM, an architect, and a technical director — but with a team that happens to be AI agents.
Planning is the real product now
Here's what I've learned from the past year: the work that takes the longest and matters the most is before any code gets written.
Defining the product. Breaking it into features. Mapping the logic. Documenting edge cases. Deciding the architecture. Choosing what NOT to build.
That's 80% of the job now. The execution — the actual coding, testing, deploying — that's the part AI handles well when you've done the planning right.
I went from a developer who spends all day coding to someone who spends most of the day thinking, defining, and reviewing. The code still gets written. The tests still get run. The product still ships. But I'm not the one typing every line anymore.
My team are agents
This isn't a metaphor. I literally work with AI agents as my execution team.
I define the task. I write the specs. I set the rules and constraints. The agent builds it. I review. I iterate. I approve or redirect.
It's the same workflow a senior engineer has with junior devs — except the feedback loop is minutes instead of days, and I never need to schedule a standup.
When the planning is solid, the output is consistent. I wrote about this before: well-planned tasks get the same result more than twice. That consistency is what makes this workflow real, not experimental.
From idea to delivery, without the time cost
The vision that excites me: take an idea — a product, a feature, a full app — and carry it from concept to shipped product. Design it. Plan it. Build it. Test it. Deploy it. Not in months with a team of 10. In weeks, maybe days, with clear planning and the right agents.
Not because the work disappears. Because the time cost shifts dramatically. The bottleneck moves from "how fast can people type code" to "how clearly can one person think."
And that's where a quote keeps echoing in my head:
"Time is the only thing you can't get back."
Every hour spent on boilerplate, on repeating patterns, on manual tasks that could be automated — that's time you don't recover. The goal isn't to work less. It's to use every hour on the part that actually matters: the decisions, the strategy, the thinking.
What I'm building toward
I don't have a title for what I'm becoming. It's not purely PM. It's not purely architect. It's not purely developer. It's someone who can:
- Take a product from zero to one
- Plan every layer — UX, frontend, backend, infrastructure
- Direct AI agents to execute the plan
- Review and iterate at speed
- Ship a complete, working product
The technical knowledge doesn't go away — it's what makes the planning possible. You can't plan what you don't understand. But the hands-on execution shifts from "I write it all" to "I design it all and review what gets built."
This is what growth looks like now
Three years ago I was learning React and figuring out how APIs work. Two years ago I was writing full-stack apps and starting to use AI for research. A year ago I was building automation pipelines and learning UI/UX.
Now? I'm planning products end-to-end and using agents to build them.
The pace is insane. What I did last year already feels like a different era. And that's not scary anymore — it's the rhythm. We either get used to it or get left behind.
I'm choosing to get used to it. Not by fighting the change, but by becoming the person who directs it.
Time doesn't come back. Use it on what matters.