The AI Doom Posts Are Getting Exhausting

I'm Actually Excited About AI

I’m getting a little tired of reading the same kind of posts every week: AI burnout is real, developers are doomed, AI is making software engineering miserable, coding is dead. Sometimes it feels like the entire conversation around AI in software has turned into a competition to see who can predict the most depressing future.

Apparently, before AI, software development was a peaceful paradise. No deadlines, no legacy code, no production incidents, no impossible tickets, no mysterious bugs at 5 PM on a Friday. We were all just happily writing perfectly documented code eight hours a day.

I don’t remember it quite like that.

The truth is that I’m actually more excited about software development now than I’ve been in a long time. Not because I think AI is magical, and definitely not because I believe we should blindly generate code and ship whatever comes out. I’m excited because AI is removing one of the biggest limitations we’ve always had as developers: the amount of time and effort required to turn an idea into something real.

For years, a lot of ideas died before they even started. Maybe you wanted to try a different language, explore another framework, experiment with a new architecture, build a small tool, investigate something outside your area of expertise, or automate an annoying part of your workflow. But then you started calculating how many evenings, weekends, documentation pages, Stack Overflow threads, and questionable blog posts from 2017 it would take, and suddenly the idea didn’t seem worth it anymore.

That barrier is getting smaller.

Today I can take an idea and explore it much faster. I can prototype something in a technology I barely know, understand how another part of the stack works, experiment with infrastructure, investigate an unfamiliar codebase, or simply test whether an idea has any value before investing days into it. That changes the economics of curiosity, and for me, that is one of the most exciting things happening in software right now.

This does not mean we should stop thinking

There is an important distinction here. Being excited about AI does not mean I believe in vibe coding everything and hoping for the best.

Experience still matters. Architecture still matters. Understanding the system matters. Reviewing code matters. Security, maintainability, performance, testing, and all the boring things we have spent years learning still matter.

AI can generate a lot of code very quickly. Conveniently, it can also generate a lot of bad code very quickly.

Our job is not suddenly to stop thinking. If anything, judgment becomes more important because the amount of output we can produce is increasing dramatically. You still need to recognize when an approach is wrong, when an abstraction is unnecessary, when a solution will create problems six months from now, or when the AI is confidently inventing something that does not exist.

The difference is that we now have much more leverage.

If something that used to take three days can now be done in a few hours, I don’t see that as three days of work disappearing. I see time becoming available for something else. Maybe that means learning another technology, improving the product, fixing technical debt, investigating a performance problem, experimenting with a better architecture, or simply moving on to another feature sooner.

That is not something I find depressing.

Technology has never stopped moving

I’ve always thought about working in tech like being on a train moving at ridiculous speed. You can enjoy the ride, but you cannot expect the scenery to stay the same forever.

Languages change. Frameworks change. Browsers change. Infrastructure changes. Deployment models change. The tools we use to write software change. Every few years, something comes along that forces us to relearn part of our job.

If you have been working in this industry for long enough, you have already done this many times.

We went from server-rendered pages to increasingly complex frontend applications. We moved from manually managing servers to cloud infrastructure and containers. JavaScript went from something people used to validate forms to something running practically everywhere. Frameworks arrived, disappeared, came back with a different logo, and occasionally reinvented PHP.

This industry has never promised stability.

AI is another one of those changes, except this one feels significantly bigger because it affects the actual process of creating software.

And yes, that means adapting again. Learning again. Being uncomfortable again.

Welcome to software development.

Code is becoming less of a barrier

One of the ideas I find most interesting is that code itself is slowly becoming less of a limitation between an idea and its implementation.

That does not mean programming languages disappear. It does not mean nobody needs to understand code anymore. But the level at which we interact with software is changing.

Instead of spending most of our energy translating ideas into syntax, we can increasingly spend more time thinking about the actual problem. What are we trying to build? What constraints do we have? How should the architecture work? What are the trade-offs? What could go wrong? Is this actually useful?

The implementation still matters, but getting to an implementation is becoming cheaper.

That opens a lot of doors.

You can dare to use another language. You can explore parts of the stack that previously felt too far outside your comfort zone. You can build tools that you would never have justified spending a week on. You can test ideas faster and throw them away when they are bad without feeling like you wasted three days implementing them.

That freedom is incredibly valuable.

For the first time in a while, it feels like imagination is becoming a bigger constraint than the mechanics of implementation.

Maybe the job is simply evolving again

I understand why some developers are worried.

There are legitimate questions around jobs, expectations, productivity pressure, code quality, and what companies will do when they realize one developer can potentially produce much more than before. Those conversations are worth having.

But I don’t think fear should be the only lens through which we look at this change.

Personally, I see an enormous opportunity to become better at the parts of engineering that go beyond typing code. Understanding systems, making decisions, designing solutions, asking better questions, identifying problems, validating ideas, and knowing when the output in front of you is complete nonsense.

Those skills are not becoming less valuable. They are becoming more important.

The way we write software is changing, and I want to understand that change instead of fighting it.

I want to learn how to use these tools properly. I want to figure out where they help, where they fail, and how they can make me a better engineer instead of just a faster typist. I want to use the extra leverage to explore technologies I would otherwise ignore, build more things, experiment more often, and spend more time understanding the systems I work with.

So no, I’m not particularly burned out by AI.

Quite the opposite.

I’m excited.

The train is moving faster again, and that has always been part of working in technology.

I’d rather learn how to drive it than spend my time complaining about the noise.