Is AI coming for DevOps / Platform engineers' jobs?

In recent times, there has been a lot of buzz about AI replacing all kinds of IT professionals. From what I've seen, the consensus is that DevOps / Platform roles are mostly safe from this for many different reasons, some of which are that companies have a lot of technical debt or legacy tooling, which makes AI integration very difficult, or that DevOps / Platform engineers require very broad permissions, which, in the hands of AI, can be very dangerous and potentially devastating for the whole company.

Personally, I do agree that DevOps / Platform engineers are in a good position when it comes to the threat of being replaced by AI. However, I think there is a different threat that is much more relevant, yet not really talked about, and that is the threat of losing your ability to think and solve problems on your own. The reason why this threat is so significant is that AI is already deeply entangled in our day-to-day jobs. Most of you probably (and hopefully) are already using AI at your job, and it has made your life easier, removing a lot of friction, making you move faster, and being more efficient.

All that is great, but at the same time, you don't have to think as much anymore; you can just dump the problem you're facing into a chatbot, review the output, maybe adjust it a little bit, and you're done. This is perfect for a lot of simple, repetitive problems or tasks, where there is not that much to learn or think about anyway, but I noticed that ever since I got this ability at my fingertips, it's usually the first thing I do, no matter if I am dealing with a mundane task, or if I am designing a new solution which involves technologies I haven't worked with before.

This, on its own, is not the problem; it's just a different, more efficient way of getting to the relevant information, just as you once would search for a GitHub issue / Stack Overflow post, or browse through a documentation. The problem lies in the fact that AI gives you a near-perfect solution which applies to your exact problem, so you can just take it, maybe polish it a bit, and you're done, while with the old ways, you still had to think about what's the best way to apply the solution to your problem.

So what can we do to remediate this? You can, of course, ask AI not to give you the complete solution, and rather nudge you in the right direction. However, this comes with a major caveat, which is that you will be much less productive than someone who doesn't do this, and hurt your work performance, which will make you lose your job much faster than losing your ability to think would.

In my opinion, the best solution to this is to accept that we, as engineers, will no longer be gaining as much new experience and evolving as quickly as we used to by just doing our jobs, and start being more proactive in gaining new experience outside of our jobs. This is, partially, why I decided to start this blog series.

As you might have guessed from the name, this blog series will be about a homelab. Homelabbing is one of the best ways to gain new experience, no matter if you are a DevOps / Platform engineer or any other engineer, because there are many different layers in the software development lifecycle, and when building a homelab, you are the one responsible for all of them, which helps you to solve many different problems you normally won't encounter during your job.

You might ask, why am I even writing this, and not just building the homelab in silence? I wanted to use this opportunity to improve my writing skills, which are also deteriorating, since I can use AI to adjust anything I write to make me look smarter, so I won't be using any AI for these blog posts. Another reason is that I am really excited about homelabbing and all the cool technology I plan to experiment with, and I would like to share my experience with others, and hopefully gain some feedback or advice.

In the next part of this series, I will explain what my plans are for the homelab, describe what hardware I chose, and why, and I will also talk more about how difficult it was to build during times when RAM and SSD prices basically quadrupled compared to a year ago, and what I did to make it as budget-friendly as possible.

That's it for now, stay tuned for part two!