Why I'm Not Concerned About Being Replaced by AI

The title of this post is a bit of clickbait, mostly because it is only partially true.

The true part is that I'm not overly concerned about being replaced by AI. But the real title should have been: "What steps am I taking to reduce the chances of being replaced by AI as a software professional?"

The truth of the matter is that I can't predict the future, so I can't really say how things will evolve. However, given the trajectory of things as I'm able to determine it at this point, I have realized that there are things I can do to reduce the chances of AI making me obsolete.

A Little Detour

Before we talk about AI, I want to take a step back and talk about something else that, in my opinion, shares similar traits.

If you had asked software professionals five or more years ago what their biggest career concern was, naturally, you would have heard completely different answers. But I believe the number one worry for many of us was age. More precisely, growing older in an industry that is notorious for being biased against mature professionals.

All of us have heard the stories: professionals in their late forties or early fifties who were laid off and then had trouble finding a new job. We have also heard about lawsuits against tech companies for age discrimination. Nothing here is new.

For me, this was not something I was willing to accept as a fact of life. Of course, I knew that there were things beyond my control. However, my core belief has always been that there is something I can do.

So I set out to understand why the software industry tends to favor the young and frown upon the mature.

Why Experience Is Not Enough

After all, when we look at lawyers and doctors, their accumulated experience often makes them more lucrative and more sought after as they age. Why isn't that also the case in the tech industry?

The answer is relatively easy to explain. Unlike many other professions, the tech sector changes at an exponential rate. The rate of change itself, the second derivative, is also accelerating. Technology that is relevant today can become outdated tomorrow. Looking only at the AI advancements of the past year, we can see this happening in plain sight.

In an ever changing industry, there is an inherent advantage to being young. Younger professionals are energized, flexible, willing to learn and accept new things, and often have much more time to dedicate to their own career growth.

Mature professionals have their experience as their main selling point. But what good is that experience in an industry that constantly changes?

Once I realized this, I understood that I had to fundamentally change the way I managed my career. If I cared, and I do, about continuously growing in my profession, I would have to get with the program.

How I Changed My Approach

For many years, I have dedicated a substantial portion of my spare time to continuing to learn and grow.

This includes reading books, articles, papers, and blog posts. It also includes continuously writing code, practicing LeetCode to improve my problem solving skills, and forcing myself to maintain this website and add relevant content through my own blog posts.

These practices do not guarantee anything in terms of my career prospects. But they make me more prepared to deal with the ever changing landscape of the tech industry, and they help keep me relevant and attractive to future employers.

Thinking from First Principles

The most fundamental change I made was in the way I think about everything.

More specifically, I realized that I needed to think about things in terms of first principles. Learning from first principles means that no matter what changes happen in technology, as long as I am able to break things down and reason about them at their core, I will be able to tackle them.

That is why I make it a point not only to learn new things, but to truly understand them from their foundations. I then implement them myself firsthand to make sure that I actually understand them.

Back to AI

At this point, hopefully, you understand why I made the detour.

The story with AI is different, but not fundamentally different. At its core, we are dealing with a new technology that is capable of performing certain tasks much faster and, to some extent, better than humans.

What worries many of us, and for good reason, is that as LLMs become better at core engineering tasks such as development and debugging, the need for engineers will decrease.

If you squint your eyes a bit, you can see how similar the two problems are. The solutions are therefore conceptually similar, with some important variations.

The first realization should be that you cannot ignore the revolution taking place in front of our eyes. You should embrace it and leverage it to its fullest. Not doing so will only hurt you. You will be left behind, and worse, ignoring AI will not save you from your fate.

On the other hand, none of us should simply accept that fate at face value. Even as this revolution takes place, there are still things we can do to remain capable, adaptable, and valuable.

The fundamental realization is that LLMs are inherently imperfect. They make mistakes, they hallucinate, and they need guidance to make good decisions.

There is another realization that came to me: the more people become dependent on LLMs, the more they risk losing their own abilities. These abilities include many of the things that made us software engineers valuable in the first place.

I have heard multiple stories about new college graduates who are incapable of explaining basic computer science concepts because they have outsourced everything to LLMs. This is exactly where we can step back and make decisions that help us remain relevant during this period.

Using AI Without Becoming Dependent on It

For example, I make it a point to code the projects I work on at home primarily by myself.

That does not mean that I don't leverage AI. But I leverage it differently, as a tutor or as a pair programmer.

AI is there to guide me when I need it, but I still do the heavy lifting.

In addition, I continue to make it a point to understand problems from first principles. I do not outsource my knowledge to AI. I use AI to teach me concepts and help me internalize them, but not to replace the process of understanding them.

Doing these things keeps me fresh, keeps me on my toes, and prepares me to take over in the situations where LLMs fall short.

Looking Forward

One thing is certain: in an ever evolving industry, with one of the biggest and most transformative revolutions already underway, remaining stagnant is simply not an option.

If we do not keep up and evolve, we will, one way or another, find ourselves left behind and irrelevant.

The price of keeping ourselves in "good shape" is not small. It requires us to remain constantly on our toes: to learn, read, write, and, most importantly, understand things from their atoms.

None of this guarantees anything about our prospects in this industry. But it makes our chances of surviving the changes much higher.

And that is what counts.