I used to be a 10x engineer. Now I'm useless.

11 March 2026 | 4 min read

Why some engineers have an existential crisis in the age of AI

I was taking a couple of weeks off from the X timeline when this reached me via a group chat on X. A post with the title “I used to be a 10x engineer. Now I’m useless.”, attached to it a 12-minute video showing an engineer explaining his experience with AI coding tools.

I can relate to Mo very well, even though I wouldn’t claim that I was a 10x engineer at any point. Maybe in pure frontend.

I think that the divide between software engineers who mourn the pre-2022 era and the “I haven’t opened an IDE in 6 months” crowd is pride in the work they did before and how they managed to do that work.

In this case (and also in mine), he was self-taught, which is an achievement in itself - one that I also take pride in.

You can be proud of yourself when you can chop down a tree using an axe and the brute strength of your body, despite the existence of chainsaws. I think Peter Steinberger recently said something along the lines of “Coding by hand will be something that people do for fun - like knitting.”

To be honest, I also see that.

I used to be a lazy person in my youth. Smart, but lazy. That changed when I started programming. Suddenly, I fell in love with doing hard things and solving (somewhat) complex problems I had never encountered before, biting my teeth out over a line of code or working deep into the night to fix a bug.

As humans, we naturally tend to take the path of least resistance, and now that is simply pushing the “easy” button and letting the AI take the wheel and build the feature or fix the bug. That leads to a disconnect between the creator and the creation, which for some is fine, and I really don’t judge here. It’s going to sound like I do, but I really get why people want to ship something ASAP and hopefully never look at it again.

Some prefer to solve a problem in any way and move on, others want to have full control and understand every intricacy of the underlying problem and its solution. I personally belong to the second camp. I like to understand systems in their entirety and to build a mental model without any gaps to solve the problem to the best of my abilities - however, it’s getting harder for me to justify the effort.

More and more, I find myself spinning up OpenCode, Codex, or Claude Code even for simple tasks, and with each time I can feel the skill - that I spent so much time and energy building - atrophy, and with it my problem-solving ability. This is why I force myself to sharpen the sword, code by hand, go through problems, and research the old-school way. Not only for the skill, but also because it’s fun, because I love it.

I’m playing with the thought of building a game by hand to scratch that itch and just move as fast as possible at my job by fully embracing the agentic route. I’ll start studying Computer Science in a couple of weeks next to my 9-5 job… we’ll see how much time I’ll realistically have to build a game, hit the gym, spend time with my family, work a 9-5, and study.

I always used to say that I didn’t care what I’d build (except something harmful) as long as I grew as an engineer and solved complex and fun problems. This hasn’t changed, and I’m more than happy to move faster through tedious boilerplate with AI, but I also like to keep some of the fun that I had in programming by doing things the hard way sometimes.

You can play a game on easy or you can play it on hard. The end will be the same, but playing on hard will give you a bigger sense of achievement, fulfilment, and you’ll undoubtedly get better at the game.

The choice is yours.