Before reading this post, ask yourself a question: When was the last time you thought really deeply?
By “thinking hard” I mean facing a specific, difficult problem and spending several days just sitting with it to overcome it.
a) All the time. b) Never. c) Somewhere in the middle.
If your answer is (a) or (b), then this post is not for you. But if, like me, your reaction is (C), you may get something from it, if only to realize that you are not alone.
First, a disclaimer: This post does not contain any answers, not even suggestions. It’s just a way of expressing what I’ve been feeling the past few months.
creator and thinker
I believe my personality is built on two primary traits:
- builder (Willingness to create, send and be practical).
- thinker (Requiring deep, long mental struggle).
The builder is pretty self-explanatory, it’s driven by velocity and usability. This is the part of me that wants to change from “thought” to “reality.” It’s like the dopamine hit of a successful deployment, the satisfaction of building systems to solve real problems, and the knowledge that someone, somewhere, is using my tool.
To explain the thinker, I have to go back to my university physics study days. From time to time, we encountered homework problems that were significantly more difficult than average. Even if you had a good grasp on the topic, it was still difficult to come up with an approach.
I noticed that students fell into three categories when faced with these problems (well, four if you count the 1% of geniuses for whom no problem was too difficult).
- Type 1: Majority. After a few tries, they give up and go to the professor or TA for help.
-
Type 2: Researcher. They went to the library in search of similar problems or insights to make the problem accessible. They usually succeeded.
-
Type 3: Thinker.
I fell into the third category, which, in my experience, was as rare as the talented 1%. My method was just to think. Think hard and long. Often for days or weeks, all my non-I/O brain time was spent constantly chewing on possible ways to solve the problem, even when I was sleeping.
This method has never failed me. I always felt that deep and long thinking was my superpower. I may not be as fast or naturally talented as the top 1%, but given enough time, I was confident I could solve anything. I felt deep satisfaction in that process.
Struggle with AI
This satisfaction is why software engineering was so gratifying to begin with. It struck the right balance. satisfied with this builder (feeling productive and practical by making useful things) and thinker (Solving really hard problems). Thinking back, the projects where I grew most as an engineer were always the ones with a large number of difficult problems that required creative solutions.
But recently, the number of times I contemplate a problem for more than a few hours has decreased significantly.
Yes, I blame AI for this.
I’m currently writing more, more complex software than ever before, yet I feel like I’m not growing at all as an engineer. When I started paying attention to why I was feeling “stuck,” I realized I was starving. thinker.
“Vibe coding” satisfies the builder. It’s great to see the idea go from idea to reality in such a short time. But it has drastically reduced the time it takes to find creative solutions to technical problems. I know many people who are complete builders, for them this era is the best thing ever. But for me, something is missing.
trap of practicality
I know what you might be thinking: “If you can ‘vibe code’ your way through this, the problem really wasn’t that hard.”
I think the point has been missed. It’s not that AI is good for hard problems, it’s just not that good for easy problems. I’m confident that my third manual rewrite of the module will be far better than anything the AI can output. But I am also a pragmatist.
If I can find a solution that is “close enough” in a fraction of the time and effort, then it is irrational not to go the AI route. And this is the real problem: I just can’t turn off my practicality.
After all, I’m a builder. I like making things. The faster I build, the better. Even if I wanted to reject AI and go back to the days where the thinker’s needs were met by coding, the builder in me would struggle with the incompetence.
Even though AI will almost certainly not produce a 100% satisfactory solution, the 70% solution it achieves usually receives a “good enough” score.
So what now?
Honestly, I don’t know. I’m still figuring it out.
I’m not sure if both parts of me can be satisfied with coding anymore. You can always aim for harder projects, hoping to find problems where AI fails completely. I still encounter them occasionally, but it seems the number of problems requiring deep creative solutions is rapidly decreasing.
I’ve tried to get a sense of mental development outside of coding. I tried to get back in touch with physics, by reading old textbooks. But that too was not successful. It’s hard to justify spending time and mental effort solving physics problems that aren’t relevant or cutting edge, when I know I can build things.
The builder side of me won’t let me sit and think about unsolved problems, and the thinker side of me is starving when I vibe-code. I am not sure whether there will ever again be a time when both needs can be met simultaneously.
“We have now the right to give it that well-known name which always signifies that no power of imagination, no flight of fearless fancy, no devoted heart, no abstract thinking no matter how deep, no enchanted and inspired soul has ever attained: God. But this original unity is of the past; it is no longer. It has broken itself, completely and utterly, by changing its being. God has died and His death was the life of the world.”
– Philip Mainlander
<a href