The Three-Body Problem of Our Industry
November 17, 2026 • Nona Dzhurkova • 6 min read

The Three-Body Problem of Our Industry

Last week, I caught myself reading only the summary.

The article was good. I was tired, my day was full, and the summary was just there. Then I thought: what did I just skip?

We live in a world where everything has to be faster. We don't have time to read, so we summarize. We don't have time to investigate, so we ask AI. We don't want to struggle, so we look for the quickest answer.

And when the answer is wrong, do we even notice?

Knowing is not the same as having access

We have more information than any generation before us. But I think we are confusing access to knowledge with actually knowing something.

Knowing takes time. You read something twice. You try it, and it breaks. You try again. You sit with a problem until it finally makes sense. That discomfort is not wasted time. It is part of learning.

I'm not against AI. Quite the opposite. I use it every day, and not just to chat. I bring it into my actual work and test what it can do and where it breaks. I want to explore everything it can give us.

But I keep coming back to one question: am I using AI to help me think, or am I getting used to letting it think for me?

What happens to our ability to learn when we outsource the effort of learning itself?

Three bodies

In physics, the two-body problem has a precise mathematical solution. Add a third body, and things become much harder. Each body pulls on the other two. Every movement changes the forces, which changes the movement again. There is no general formula that can predict every possible outcome.

A small difference at the beginning can lead to a completely different result. In some configurations, one body can even be thrown out of the system forever.

This is our industry right now.

The first body is the junior. They studied for years. Then they hear: we want experience. But nobody gives them the first chance.

Some do get hired, and receive documentation, an AI tool and a list of tasks. Nobody explains why things work the way they do. And when they struggle, we start doubting them.

The second body is the senior. That's us. We have the experience, the scars from production at 2 a.m., the feeling that tells us something is wrong even when everything looks right. But we are busy, and mentoring becomes something we'll do when there is time.

And some of us have become comfortable. We think what we know is enough.

It isn't.

The third body is AI. It is fast, it never gets tired, and it is quietly taking over some of the small tasks, the same tasks juniors used to learn from.

Now look at how they influence each other.

When AI gets stronger, it can pull the junior away from the senior. Why ask a person when the answer is instant? The senior sees fewer questions and assumes less support is needed. They drift apart, and each one thinks the other chose it.

When the junior relies only on AI, they get answers without necessarily understanding them. They may not even know what to question. And when AI is wrong, nobody notices.

But there is something even bigger at stake. What happens when the people who should become tomorrow's seniors never get the chance to build the experience we now take for granted? Who will mentor the next generation?

Like in physics, small things at the beginning can change everything. One manager who says, "Let me show you why." One company that takes a chance on a junior. One senior who stays curious and makes time to teach.

These things may look small. They aren't.

Here is the part that gives me hope. The three-body problem doesn't mean stable solutions are impossible. There are special configurations where three bodies can move together in stable patterns. Stability is possible, but it doesn't happen by accident.

I believe mentoring is part of that force. It helps keep all three in orbit, so nobody gets thrown out into the dark.

What I don't know

I don't know what software development will look like in five years. I don't think anyone does. That is what makes it a three-body problem: we cannot predict every outcome, but we can still influence what happens next.

But I know what is true today.

Producing code is easier now. Finding information is easier now. Getting an answer is easier now.

Understanding is not. And I worry that we are starting to lose it.

What we can do

If you are a senior: teach. Explain the why, not just the fix. Someone once gave you a chance. Be that person for somebody else.

But don't stop learning yourself. Sometimes a junior's question makes you reconsider something you've always taken for granted. Experience gives you something to share. It doesn't mean you are finished learning.

If you are somewhere in the middle: stay curious, even when nobody asks you to. I don't believe anyone should have to give their evenings to work. Life is bigger than a job. But curiosity cannot depend only on what happens to be in the backlog. The industry won't wait for us.

If you are trying to enter this industry: please don't let rejection convince you that you don't belong here. Keep learning. Build things. Break them. Ask questions. Use AI, but take the time to understand what it gives you. Find people who are willing to share what they know. Your first opportunity may take longer than it should. Don't give up on becoming really good at what you do.

The part that stays human

We don't just need people who can generate code. We need people who can question an answer, recognize a problem, connect ideas, make decisions and understand the consequences.

AI can give the answer. But it cannot replace the experience of someone sitting next to a nervous junior and saying, "I believe in you. Let me show you why."

The future may belong to AI. But the responsibility to think, learn and grow, and to take someone with us, still belongs to us.

If you read this until the end, thank you. You just did the thing I'm afraid we are forgetting how to do.

Featured Blogs