It’s 2026, and I’m writing this post without AI. That already feels like a confession.
In the span of a year, AI has fundamentally changed how I work as a software engineer, architect, and consultant. We’ve gone from hallucination-prone Q&A toys to agents that can plan, debug, and implement features - largely on their own.
This shift has led to many outcomes. I spend most of my time:
- Thinking about solving problems, planning alongside an AI agent.
- Asking AI agents to execute these plans
- Reviewing its output and iterating on improvements
On the whole, when using frontier models such as Claude Opus & OpenAI GPT-Codex it does a fantastic job of planning features, and implementing them. Success varies on how well the problem domain is understood, and how good the context you provide it is. The more you put in, the more you get out.
Whilst a lot of people are focusing on the discussion around what this means for the job market and whether it will lead to lay-offs, I’ve mainly feeling a little sad about what it means for the job itself.
Many people assuming I meant job loss anxiety but that's just one presentation. I'm seeing near-manic episodes triggered by watching software shift from scarce to abundant. Compulsive behaviors around agent usage. Dissociative awe at the temporal compression of change. It's not… https://t.co/UkdjxtALut
— Tom Dale (@tomdale) February 6, 2026
Tom Dale beautifully put some of this into words this week when he wrote about the “Dissociative awe at the temporal compression of change”, for those who are completely overwhelmed at the speed at which our role is changing.
For me, it’s not just about the speed at which things are changing but what it means for the job itself. The key component of our profession.
Programming, at its core, has always been about building.
The joy of architecture. The joy of design. The quiet focus of shaping an abstraction until it feels right. The satisfaction of watching a messy system become elegant. The small euphoria when the tests finally turn green.
I can trace that feeling all the way back to building Lego as a child. The quiet concentration, the click of the final piece snapping into place, the simple pride of having made something.
The joy only works if you’re the one putting the pieces together.
And that’s what feels different now.
With AI doing the heavy lifting, I’m not building in the same way. I’m supervising. I’m coordinating. I’m reviewing and approving work produced at a pace no human team could match.
The agents don’t just assist; they execute. They don’t just suggest; they implement. In a day, they can do what once took a team a month.
You can’t put the cat back in the bag. The capability is here. So what does that make us?
Do experienced engineers want to become coordinators of autonomous systems? Gatekeepers for swarms of agents? Bottlenecks responsible for sanity-checking the output of something that works 100x faster than we ever could?
Maybe some do. Maybe that becomes the new craft.
But I can’t ignore the quiet sadness that comes with it.
I’m in awe of what AI can do. Truly. It’s one of the most extraordinary technological shifts I’ve witnessed.
And yet, the part of the job I loved most - the act of building, of wrestling with code until it yielded, of snapping the final piece into place - feels like it’s slipping out of my hands.
I don’t know where this ends.
I only know that something fundamental has changed.