The AI Fog-of-War on Software Engineering Jobs
There’s a huge fog-of-war surrounding the impact of AI on software engineering jobs. Everyone has an opinion. Half the internet thinks engineers are six months from unemployment. The other half thinks nothing will change. Both are wrong. Let me clear some things up.
Fundamentals Matter More Than Ever
Having knowledge of computer science fundamentals is more important than ever, and that is a direct result of AI automating the “grunt work.”
Think about what that means. The rote stuff — boilerplate code, standard CRUD operations, wiring up APIs — is getting handled by LLMs at an accelerating rate. What’s left is the hard stuff. The architecture decisions. The performance trade-offs. The “this answer looks right but will fall apart at scale” judgment calls. You can’t make those calls if you don’t understand the fundamentals.
Software engineers need to know when to call an LLM out on its BS, and that is impossible to do unless you understand what “correct” looks like. An LLM will confidently hand you a solution that compiles, passes basic tests, and quietly introduces a race condition or a memory leak that won’t surface until production. If you can’t spot that, you’re not engineering. You’re copy-pasting with extra steps.
Advice for Junior Engineers
For incoming junior engineers, you won’t have the experience that allows seniors to catch those mistakes through pattern recognition and battle scars. But you can have something even more valuable: understanding how electricity moving through a wire turns into pixels on a screen.
That sounds dramatic, but I mean it literally. Make sure you graduate university with an understanding of compilers, operating systems, CPU architecture, networking (the OSI model), and everything in between that facilitates the process of turning electricity into light waves. That full-stack mental model — from silicon to screen — is what separates engineers who can reason about systems from engineers who can only use frameworks.
When an LLM gives you an answer and you understand the entire chain of what’s actually happening under the hood, you’re dangerous in the best possible way. You can leverage AI as a multiplier on top of real understanding instead of using it as a crutch to mask the absence of it.
The Imposter Syndrome Risk
Ever feel like an imposter and that there is too much to possibly know about CS fundamentals but you’ve somehow been hacking it as a software engineer?
I’m not going to sugarcoat this: you’re at extreme risk. You’re at risk of being replaced by a software engineer who has knowledge of the fundamentals and uses AI dev tools. These are the people who previously would have been fast-tracked into engineering management because their deep understanding made them too valuable to leave in an IC role. But now businesses are figuring out that they can have an entire layer of these people actually writing code — because AI handles the tedium that used to slow them down — and they’re shifting their budgets accordingly.
One engineer with strong fundamentals and modern AI tooling can now do the work that used to require a small team. Companies have noticed. Headcount is being re-evaluated everywhere.
AI Will Not Take Your Job
AI will not take your job. Someone using AI will.
If you’re currently in a CS program: you’re in the right place. I promise, the future of the industry is very bright. The demand for people who actually understand computing is going up, not down. Keep going.
If you’re currently a software engineer worried about being displaced by the revolution happening in AI: fill in your knowledge of the fundamentals and use AI to help you learn. Seriously — the irony is that AI is one of the best tools ever created for learning the very things that make you irreplaceable by AI. Use it.
The End of the Passion-Free Era
Last, but not least, the era of employing software engineers that do not have a passion for technology is over.
There was an odd phenomenon over the past decade where people could rush through a bootcamp and land a six-figure job without actually being passionate about technology. The pitch was simple: “learn to code, make money.” And for a while it worked, because demand was so insatiable that companies would hire anyone who could produce a functioning React app.
Those days are over and I doubt they will ever come back. When AI can produce that same functioning React app in minutes, the bar for what a human engineer needs to bring to the table goes way up. You need curiosity. You need to actually care about how things work. You need to be the kind of person who reads about new technologies on a Saturday morning because you find it genuinely interesting, not because your job requires it.
Good riddance.