Self-Knowledge with AI: The Greatest Investment a Developer Can Make

How a journal, a to-do list and an honest conversation with AI changed the way I see myself and the way I build software.

In my technical accounting course back in the ’90s, there was a class on psychology. Out of everything I learned that semester, one phrase stayed with me until today: “Know thyself,” inscribed at the Temple of Apollo in Delphi and made eternal by Socrates.

Twenty-something years later, with a full career in software engineering behind me, I have realized this is the most underrated skill in our profession. We collect certificates, frameworks, languages and architecture. We stack courses on LinkedIn as if they were trophies. But we rarely stop to debug ourselves.

And what an irony. We, who spend our lives fixing other people’s bugs, are the ones least likely to look at our own.

Photo by Melissa Holowaty on Unsplash

Why This Matters for Those Who Write Code

Socrates also said that “an unexamined life is not worth living.” I would adapt this for our craft: a career without reflection is not worth building.

How many times have you accepted a project knowing it would drain your energy? How many times have you fought over a technical decision that, deep down, was just ego? How many times have you postponed rest because “just one more commit”? These are all symptoms, and symptoms must be treated at the cause, not in the stack trace.

The cause is almost always in a place that hurts to look at. That is why most people do not look.

The Method That Works for Me

There is no shortcut to self-knowledge, but there are tools. Mine are absurdly simple.

1. A Journal Connected to Goals

This is not a teenager’s diary. It is an engineer’s log. I write down epiphanies, decisions, situations that pushed me over the edge, and patterns that keep repeating. Will Durant said, “We are what we repeatedly do,” so I started observing what I was repeating.

2. Reflection Before Sleep

I replay the day in my head and imagine a better version of myself reacting to the same situations. It sounds silly, but it is mental training, just like writing tests for code that does not exist yet. You are programming future behavior.

3. Microsoft TO DO as a Measurement Instrument

I record everything I do during the month and categorize it. The categorized data becomes the input I feed to the AI later (more on that below).

Life Happens in Cycles

One thing I’ve noticed from keeping this habit is that life repeats itself in cycles, and the same lesson keeps coming back until you learn it. Every day, trivial situations (or not) will reappear. If you have already mapped out a better way to react in your head, you will change automatically when they return.

And there are moments when the entire system shifts: a trip, houseguests, a new project. When the ecosystem changes, a window opens to rewrite behaviors. It is like deploying a new version. The environment is already in maintenance mode, so take advantage of it.

I use these windows consciously. But I can only do that because I already know what I want to change. And that knowledge only came from the habit of observing myself.

Where AI Comes In, and Where It Falls Short

Part I: Using the TO DO

I use the TO DO tasks at the end of the month, export them and feed them to AI, asking for an honest analysis: what made me waste time, what I should eliminate and where I am running away from my own goals.

I did this throughout 2025 and into early 2026, and the results scared me, in a good way: too many tasks for a single developer; too many projects in the timeline; and it helped give me focus on the main solutions.

Part II: Using the LLM

This is the part that helped me the most over the past two years. The exercise goes like this.

1. Create a long document with the story of your life. Be detailed. List your traumas, your victories, your shames, your patterns. Be brutally honest.

2. Add a section about what you want to learn in the coming months.

3. Add a section about what you want to achieve in life.

4. Then ask the AI model:

“What did I fail to notice in my own story that is hidden in what I wrote?”

The result may surprise you. It surprised me. AI sees patterns you do not see, because you are inside the pattern.

But, and this “but” is large, AI hallucinates. It will make things up. It will reach wrong conclusions. It will tell you what you want to hear if you give it room. Use the result as a hypothesis, never as a diagnosis. The ideal is still a real therapist.

I use AI as a complementary mirror, and when something really hits me, I take it to someone who knows me, a close friend or a family member, to check it against a human perspective.

AI is a tool. A good and powerful one, but a tool. It does not replace therapy. It does not replace friendship. And it does not replace the courage to look at yourself.

Why It Hurts and Why It Is Worth It

Knowing yourself hurts. You may face things you would rather have forgotten. But those are exactly the things blocking your progress.

Nobody can help you out of your own shell. There is a cruel observation in nature: if you help a chick break its shell, it dies. It did not develop the strength needed for what comes next. Our shell works the same way. The effort itself is the development.

Seek adversity on purpose. Look for challenges that scare you a little. Try to be better than you were yesterday, not better than the developer at the next desk. You are your own obstacle, and nobody is going to fight that fight for you.

Conclusion

This is the path that worked for me. It will not work the same way for you. You need to find your own tools, your own pace, your own style of honesty.

But start. Open a document today, write your story, write what you want from life and what is holding you back. Reread it in a month. Reread it in a year. You will see things that are invisible to you today.

Investing in new frameworks gives you a few months’ advantage. Investing in yourself earns you a lifetime.