Happy April.

Don't forget to comment, like and subscribe :)

Social Media:
Follow me on GitHub: https://github.com/Elfocrash
Follow me on Twitter: https://twitter.com/nickchapsas
Connect on LinkedIn: https://www.linkedin.com/in/nick-chapsas

Nick Chapsas

A narrative journey through a software engineer's life from childhood curiosity to senior engineer, culminating in an existential crisis triggered by AI tools replacing the skills and identity they spent decades building. The story traces the arc from first touching a computer, through university, early career struggles, a failed Microsoft interview, eventual seniority, and finally the unsettling realization that AI can now do in minutes what once defined them. It ends with a deeply personal question about identity when your craft is automated away.

Why It Sucks To Be Born as a Software Engineer

no harsh feelings but I’m pretty certain writing code was not the part of my job. It was designing the architecture, the code was just there because we needed the architecture to involve a computer, and also because humans can’t read 0s and 1s like a machine. But seriously though, it’s not much of a problem AI actually unmasked bad programmers and only made the job of good programmers harder, so it’s not really a big problem… totally not a big problem.

I feel your pain bro, I loved writing. This was core of my existence. Now I don’t know who I’m. Overseeing AI is opposite what I love to do

I don’t feel threatened by AI. I think people who are skilled engineers who are actually or somewhat fearful may be unwittingly buying into the hype train a little too much.
It definitely raises the bar and is unquestionably a useful tool, but it often gets basics at least partially wrong, and I find that it really struggles with actual engineering problems. Sure, it does ok with well-defined or common tasks, but I find LLMs still frequently produces bad or just outright incorrect information.
Earlier today I had GPT 5.4 -xhigh via Codex tell me a problem existed that didn’t actually exist, despite the fact that had nothing to do with what I had instructed it to do, which was to write several DTO’s based off an OpenAPI spec (which is simple task, but very tedious). Had I allowed it to “fix” the “problem” it insisted existed, it would’ve actually broken an unrelated (to it’s task) part of the codebase.
LLM’s have also had a terrible track-record IME when it comes to SpringBoot (I’ve found they often confuse major spring versions, even when you tell them what versions of SpringBoot you’re using, if they even give you a correct answer), Reactive programming in general, or really anything that isn’t or hasn’t recently been fairly popular in the mainstream.

AI is still a tool that is based on probability and that’s why it is still not what people or AI CEO’s claim it to be. Many people bite the AI narrative simply because they haven’t got the knowledge software engineers have. I believe the hype will soon die down and the market will naturally correct itself.

I feel lucky no matter how it goes. It was a good journey and it’s still not over.