I’ve been trying to put my finger on what feels fundamentally different lately. Not the obvious stuff…yes, we have AI assistants now. But, it’s deeper than a new monthly subscription.

We’re at this odd inflection point where each of us has access to a coding machine that is, in most practical senses, the best coder in any room we walk into. This isn’t hyperbole. That’s just a truth we have to live with now.

So, what happens when the baseline technical bar gets moved by something that isn’t human?

The Ladder Just Got Taller (Or did we pull it up?)

For most of my career, the progression felt predictable. Junior devs could write working code but needed guidance. Mid-level devs could ship features independently. Seniors could architect systems and see around corners.

But the real differentiators as you moved up weren’t just about raw coding ability and systems design. The qualities that got you promoted were softer, and developed from years of scars, not theories: the ability to zoom out and see the macro while still managing the micro. Communication. Processes that worked for for humans, not just machines. Managing other developers, making the team better.

Those were senior skills precisely because in most people they needed years to develop.

Now? The barrier to entry is suddenly higher because AI handles what used to be the early-career learning curve. If a junior can generate decent code with the right prompts, what are they actually learning in those first two years? If they can do that, the differentiators become those same soft skills that once defined seniority.

The job market reflects this in ways that are uncomfortable to talk about. Companies are hiring fewer juniors. The ones they do hire are expected to already think like seniors. Add on top the entry level position that requires 2-3 years experience.

The Cruel Irony

Despite everyone’s best efforts to replace us, someone still needs to run the code machine. Someone needs to understand what the AI generated, why it made those choices, and whether it’s actually solving the right problem. Someone needs to maintain all this generated code, refactor it, and, yes, debug it when it inevitably breaks in production.

And suddenly junior developers are expected to be project managers and requirements gatherers, and expert communicators from the beginning as those are the skills required to run the code machine.

Now the job is basically: manage the AI, translate between stakeholders, explain technical decisions to non-technical people, and coordinate work across teams. Oh, and also maintain the mountain of generated code that nobody fully understands because it was written in 30 seconds by an LLM.

It’s like we’ve become middle managers for machines. The human bottleneck remains, just moved up the stack.

Craftsmanship vs. Production

This raises a personal question that I think we’re all grappling with: where do you find enjoyment in your day now? What gives you that spark or sense of accomplishment?

For me, coding has primarily been a means to an end. I like building things people can use. The actual typing-code-into-editor part was never the point, it was just the most effective tool I had. So while I love coding in a professional context these shifts don’t bother me as much. If anything, AI makes the means-to-end faster.

But I know people who devoted their lives to the craft of programming. People who loved the puzzle-solving aspect, the moment of elation when you finally crack a complex algorithm. That dopamine hit of making something elegant work.

That… just doesnt feel the same when it is AI producing the code. It’s transformed. A production mindset has taken over: ship faster, iterate more, generate and test and deploy. The craftsmanship feels less central. I don’t think that’s necessarily bad. But I imagine it lands differently depending on what drew you to this work in the first place.

The Human Computers

I keep thinking about mathematicians and what happened when computers arrived.

Before digital computers, complex calculations still needed to happen. Finance, engineering, physics, everything required computation. We had human computers, people whose job was to perform calculations by hand. They’d spend days working through differential equations or astronomical tables, anything really.

Mathematicians created the machines that eliminated the need for human computers. The parallels feel obvious.

But here’s what would’ve shocked anyone at the time: mathematics didn’t die. It exploded.

The focus shifted away from tedious calculation toward conceptual understanding. New fields emerged. Things that were literally impossible to explore when every calculation took weeks. The field became more important, more creative, and more powerful.

Gone are the human calculators. Nobody needs them anymore we all have access to a personal calculator. I think about this a lot when people ask if AI is replacing developers.

The Wrong Question (And What Keeps Us Awake)

“Are developers being replaced?” is not the interesting one.

Better questions: How many developers do we actually need? What will our job function be? How do we mentor and raise the next generation when the traditional path doesn’t exist anymore? How do I fit into this future? Is this what I signed up for?

We learned by doing. Years of making mistakes, debugging awful production issues, slowly building intuition about systems.

If AI short-circuits that process, where do the next seniors come from? You can’t just prompt-engineer your way to architectural judgment. Some kinds of knowledge still require experience, failure, repetition.

Maybe we figure out new paths. Maybe we’re already figuring them out and just haven’t named them yet. I don’t know. I’m honestly uncertain here, and anyone who claims to have this figured out is probably selling a subscription.

Agency Over Efficiency

As you position yourself for whatever comes next, here’s what feels true to me: don’t become the human computer.

The human computers were excellent at what they did and they were also completely replaceable the moment a better tool came along.

The mathematicians who created the computers? They’re fine. More than fine. Because they understood the concepts, the systems, the why behind the work.

Having agency, understanding not just how to use the tools but when and why, being able to design systems and make judgment calls and talk to humans, that’s more critical than ever. The technical skills are still important. But they’re no longer sufficient on their own.

Which is maybe how it should’ve been all along. We just had a few decades where pure technical ability could carry you pretty far, and that window might be closing.

My Stakes

I should probably end with something definitive and perhaphs inspirational. The general observations about the industry are over; now, the personal frustration/relief.

For most of my career, I’ve felt like my ideas were bigger than my skills or my attention span. I’d sketch out some ambitious idea, but then get bogged down in implementation details, or lose interest, or just fail to communicate the vision clearly enough.

For me, AI is genuinely empowering. It helps me ideate, structure my thinking, and rapidly produce proof-of-concepts that I can actually show people instead of vaguely gesturing at whiteboards. The gap between idea and working prototype has collapsed in a way that feels almost magical.

I’m aware this makes me an optimist in a sea of anxiety. Maybe I’m naive. Maybe I’m not seeing the full picture (how could I working for a smaller company).

But I figure in situations like this, it’s best to just charge forth enthusiastically until we meet our untimely and tragically predictable demise. At least we’ll have built some interesting things along the way.

And if nothing else, we won’t be the human computers.