Amidst the AI Hype

I've always loved writing. It's one of the few things that forces me to slow down. To take time and put words in an order that someone else can understand. In a time where I feel like I'm constantly ingesting information faster than I can process it, forcing myself to write is my attempt to connect with my humanity again.

So I return to it, partly as a reset and partly to prove to myself that I can still think clearly without relying on AI to fill in the gaps. When the entire world is flooded with AI-generated content, knowing your mind seems harder than it's supposed to. This is my attempt at an answer.

AI Doom and Gloom

As I search for a topic to write about, I keep coming back to the same thing: the state of software engineering right now, and what it means to grow through the noise as a junior developer. I'm not writing from the perspective of a seasoned professional. I'm just a fresh college grad with less than one year of experience. Every day, I wake up to a flood of doom and gloom: headlines claiming software engineering is coming to an end, articles about layoffs, LinkedIn posts arguing there's no point in learning to code anymore. To be frank, the trends I'm seeing across the industry do reflect some truth in these claims. But I've come to believe that sitting with anxiety does no good, and that only by unpacking my own worries can I make sense of what to do next.

What I'm Seeing

From where I stand, the changes are subtle but real. The most visible shift isn't that AI is replacing engineers wholesale. It's that the texture of the work is changing. Tasks that used to take weeks can now be done in days. The feedback loop between having an idea and having working code has compressed dramatically. Emphasis on "working", not "correct" code.

On one hand, the speed at which I can execute on a vision feels genuinely exciting. On the other, it's quite disorienting. When you're just starting out, confidence is built by struggling through the slow, frustrating grunt work of figuring out how different components connect, drawing mental models of the system you're working in, and failing enough times to internalize something in your own understanding. That struggle is being short-circuited now, and I don't think it's going to slow down.

Even so, I don't think this means AI tools are bad, or that junior engineers should avoid using them. But there's a difference between using them thoughtfully as you're trying to hone your craft and outsourcing your thinking entirely. Learning to navigate that line early seems more important than ever, and you rarely hear people say this out loud.

How I See This Play Out

It will likely get worse before it gets better. As AI continues to disrupt and change how work is done, it will take time for companies to re-organize, either through mass layoffs, re-investments, or re-training their workforce. The uncomfortable truth is that AI is moving faster than most organizations know how to absorb, and the gap between what AI tooling can do vs. what teams know how to responsibly use them are still enormous.

Looking ahead, I remain deeply skeptical of predictions that land in either extreme. No, I don't believe AI will bring about a techno-utopia where cancer is cured and humans spend their days maximizing happiness and creativity. I also don't think Skynet happens, and that AI brings about the end of humanity. The people claiming software engineering is over aren't accounting for how much demand for well-built software still exists. And the people dismissing all concern as overblown hype aren't paying close enough attention to how quickly entry-level work is being restructured. My best guess is that we land somewhere in between.

History offers some reassurance here. When the loom automated textile work, when the assembly line restructured manufacturing, when the spreadsheet made entire accounting roles redundant, each wave brought forth economic disruption. Jobs disappeared. Industries contracted. People worried, reasonably, that there was no place for them in what came next. But in each case, businesses eventually re-oriented around the new tools, and job categories emerged that nobody had thought to predict beforehand. To put it simply, the loom didn't end work, it changed what work meant.

AI will likely follow a similar pattern.

Work, Identity, and the Uncomfortable Truth

There is, however, something else changing beneath the surface that I think people feel subconsciously but struggle to put into words. For a lot of us, what we do and who we are got tangled up somewhere along the way. After all, it's hard not to build your identity around your craft when a large majority of your adult days are spent practicing it. The problem is that the craft we tied ourselves to happens to be exactly what AI agents are getting good at.

To quote the great Michael Novati himself: "For two centuries, modern economies rewarded people for performing scarce cognitive labor, namely calculations, analysis, pattern recognition, technical fluency. These were treated as the highest expression of talent because markets valued them most. But machines are becoming extraordinarily good at many of those same tasks. And when that happens, something uncomfortable becomes visible: some of what we called 'talent' was actually temporary scarcity."

For someone early in their career, that's not easy to sit with. I remember choosing this field partly because of what it seemed to promise — a craft that rewarded deep thinking, one where getting good at something meant something. That promise hasn't disappeared, but it's been complicated. And I think it's worth being honest about that, rather than papering over it with optimism.

What to Focus On

So, where does that leave us? How do we move forward amidst the noise, and what can you do if you're just starting out in this field? There are 3 things I think worth paying attention to:

• Build on fundamentals. The hype cycle is loud, and if you're not careful it'll convince you that the only things worth learning are whatever tool is popular this month. But I'd argue that the fundamentals — how systems actually work, how data moves, how to reason and debug — matter more, not less. The path to building good judgement and instinct lies in having a strong foundation to stand on. On this axis, I recommend picking a niche and diving deep. Build projects on your own time, find something you're genuinely excited about, and use AI to guide you through implementation. But occasionally put the tools down and write code yourself. Reason through the edge cases, sit with the confusion, let the understanding settle in your own mind. Don't outsource your thinking. Take the steering wheel instead.

• Lean into AI tools, but stay in the driver seat. I believe the future of software engineering will involve some version of AI agents taking over the grunt work of coding, either through pair programming (prompt-feedback cycle) or autonomous sessions where a swarm of AI agents operate in parallel to execute on a spec doc. But here's the thing: someone still has to know where to go. In order to effectively direct these tools, you need to understand the problem space well enough to recognize when the output is wrong, incomplete, or missing the point entirely.

• Develop taste, direction, and purpose. This is the one I keep coming back to. AI can generate code, but it can't tell you what's worth building, or why, or how to make the hundred small judgment calls that add up to something coherent and useful. We're ultimately still building for humans, not AI, which means understanding the person on the other end of what you're building still matters more than understanding the tool you're building it with. Taste is what lets you look at two working solutions and know which one is better. Direction is what keeps you from chasing every new thing. Purpose is what gets you through the stretches where the work is hard and the progress is slow. In a time where everything feels uncertain, I'd argue that knowing why you're here and why you love to build software is the best thing you can hold onto.

At the End of the Day

I know none of this is easy to sit with, and I won't pretend I have it figured out. I'm not sure anyone does, but I've made my peace with that. What I do know is that the engineers who thrive through whatever comes next aren't necessarily the ones who keep chasing the next shiny thing. They're the ones who stayed curious, kept working on their foundations and weaknesses, and didn't let the noise erode their sense of why they started in the first place.

If you're reading this and you're in a similar place — early career, a little overwhelmed, and trying to figure out what to do next — I hope this was worth your time. I know I can't offer much in the way of certainty, but I can offer this: you're not alone in feeling the weight of it.

We're going to be okay.