Expertise Is No Longer the Main Barrier

Over the last few days, I have been going through a small reassessment. I watched several podcasts, reconsidered my own experience, and arrived at a fairly simple thought.

Today a person can enter almost any field, even if at the beginning they know almost nothing about it. But this requires two abilities. The first is the ability to design systems. The second, and even more important one, is long-term motivation and the ability to bring those systems to the point where they actually work.

This does not mean knowledge is no longer needed. It means you no longer have to spend years accumulating all the expertise inside your own head before you begin. You can build a system around a task from AI agents, research, data, and checks. Your job is to set the right goal, divide it into parts, connect the results, and notice when the system loses contact with reality.

Over the last year, I tested a huge number of market hypotheses. The funny thing is that I found almost no idea that someone had not already studied before me. For every idea, even the craziest one, there were already papers from universities, laboratories, or funds. This changed my view of my own expertise.

After nine years in trading, I suddenly understood how little I knew. I previously paid almost no attention to the market through options. Now I see that they provide a completely different view of risk, the expectations of large players, and the construction of a trade itself. And the deeper I go, the more often I ask a neural network to explain something more simply, with basic logic and analogies.

Expertise sometimes helps, and sometimes locks a person inside overly rigid beliefs. A newcomer can ask a question that a specialist stopped asking a long time ago. But turning that question into a working product still requires a verification system and a very long commitment.

I like to think that today almost anyone can become the manager of their own small corporation. They can have tens or hundreds of AI agents researching, calculating, checking, and assembling separate parts. The human still has to see which system is worth building and avoid abandoning it after the first ten dead ends.

So I think less and less: "Am I expert enough to work on this?" And more often: "Can I build the system correctly and keep pushing it long enough to produce a result?"

That shift is genuinely breathtaking.