The Greenest GitHub Era and the Quiet Loss of Craft
Builders are building 10x, shipping 50x, and commit counts have exploded 100x. Everyone’s GitHub graph looks greener than a rainforest after a storm.
Features that once took weeks now appear in days.Entire prototypes materialize from prompts.
As an engineer, I should be thrilled. And part of me genuinely is. Watching AI evolve at this speed is extraordinary. It feels historic like we’re witnessing a technological inflection point that future generations will read about.
But if I’m honest, there’s another feeling underneath the excitement. A quiet discomfort. A slow grief.
Alongside all this speed, something else feels like it’s fading.
I used to take pride in the craft of engineering. Not just in shipping, but in understanding. In struggling with a bug for hours, trying different approaches, tracing logs, re-reading documentation, testing assumptions, and finally…finally, seeing the solution click into place. That moment wasn’t just relief; it was earned satisfaction. It was growth.
Solving something difficult reshaped how I thought. Writing about it afterwards felt meaningful.
The dopamine wasn’t from the praise, it was from knowing I had wrestled with complexity and won.
Productivity x AI
Today, we are undeniably more productive. Code gets generated instantly. Patterns are suggested before we consciously articulate them. Entire pull requests can be scaffolded in minutes.
And yet, I haven’t met many engineers who feel lighter because of it.
Work-life balance hasn’t magically improved. In many cases, expectations have simply scaled with productivity. If we can ship 10x faster, we are expected to ship 10x more.
If AI can generate code instantly, we’re now responsible for reviewing, validating, correcting, and cleaning up just as quickly.
A different kind of fatigue has entered the room not from writing everything ourselves, but from supervising an endless stream of generated output.
The common response to this concern is predictable: “This is where experienced developers shine. Learn to work with agents. Adapt.”
And I understand that argument. Truly. There is value in orchestration. There is skill in guiding systems.
AI babysitting
But sometimes I don’t want to orchestrate.
Sometimes I don’t want to argue with a model about intent, or trace through code that technically works but lacks cohesion or thoughtfulness. Sometimes I don’t want my role to feel like AI babysitting.
I want to build with humans. I want to collaborate, debate architecture, challenge ideas, and think deeply about trade-offs. I want the joy of shaping something from the inside out, not just validating what was auto-generated.
It’s also worth noticing who seems most excited about this era.
- Founders
- Leadership teams
- Business owners
- Indie hackers
- Non-technical operators who can now build products without years of training.
And that’s not a criticism. AI is lowering barriers. It’s democratizing creation. It’s unlocking speed and access in ways that are genuinely powerful.
But I’m not a founder. I’m not chasing venture scale. I’m not building a solo SaaS empire.
I’m an engineer.
Cry of an mediocre engineer
I worked hard, very hard to develop my skills. I invested years into understanding systems, patterns, trade-offs, performance, architecture. And now the definition of value feels like it’s shifting faster than I can process.
In a few years, the gap between what agents can do and what senior engineers can do will narrow. Maybe not completely, but enough to change hiring strategies and team structures. Fewer junior roles may exist. Mid-level responsibilities may evolve. Senior engineers might become primarily reviewers, designers, or AI supervisors.
Will there be new roles? Probably. History suggests there will be.
Engineering has reinvented itself many times before from assembly to high-level languages, from on-premise infrastructure to cloud-native systems.
But transitions are rarely gentle.
There will likely be job losses. Roles will consolidate. Some engineers may decide the uncertainty isn’t worth it and leave the industry entirely. Others will adapt and thrive. Many will sit somewhere in between anxious, tired, unsure.
And beneath all of this sits a deeper truth: this isn’t only about productivity or job security. It’s about identity. AI doesn’t just accelerate output. It compresses the journey.
Is AI Wrong?
That doesn’t mean AI is wrong. It doesn’t mean innovation should stop. It doesn’t mean the future is doomed p(doom)
But it does mean we’re allowed to feel conflicted.
We can be impressed and unsettled at the same time. We can use the tools and still question what they’re changing in us. We can adapt while also grieving the version of engineering we fell in love with.
Hope is a good thing
I don’t hope AI fails. I don’t hope progress slows down. What I hope is simpler:
that we don’t lose the human dimension of building. That engineering remains more than prompt engineering. That curiosity, depth, and thoughtful design continue to matter not just velocity metrics and commit graphs.
Because building was never only about shipping faster. It was about learning, understanding, and creating something that reflected our thinking.
If the future of engineering still makes room for that for depth, for mastery, for human collaboration then maybe this transition will lead somewhere meaningful.
But if we reduce engineering to pure speed and output optimization, something quietly beautiful will have been lost.
And it’s okay to say that out loud.