50% of people using AI just open a browser and treat it like a search engine to ask questions; 30% know how to assign tasks based on the model's characteristics; 15% can bring their own experience into AI and establish feedback loops; only 5% will delve deeper into the model.

But among that 5%, more than half spend their time building so-called "dashboards." Honestly, that doesn't help your decision-making at all—you're just making something that looks cool. Only a very small number of people can use AI to build projects that generate continuous profits, and their profit formulas are almost always based on "information asymmetry."

What is information asymmetry?

First, time lag. Every morning when I wake up, I spend 15 minutes reading the latest overseas information. On average, 6 hours later, you'll see media KOLs publishing content I already knew to harvest traffic; a few hours after that, local financial media start reporting. This information gap of up to 12 hours precisely reflects the lack of innovation in traditional media. Imagine how much you could do with that 12-hour advantage.

Second, the cognitive gap in financial rules. From options volatility to the underlying rules of US stocks, I believe most people have only a fuzzy understanding. Just understanding the mechanism differences between different brokers (including local sub-brokerage) and cryptocurrency exchanges is enough to create a massive information gap. For example, the correlation between Korean stock market openings, US stock freeze mechanisms, and index tracking—these data points are enough for me to predict the opening gap direction of US stocks on Monday.

Third, actively break out of your echo chamber. You must start the filtering exercise of "I'm not interested in this post." When you see those posts that start with exaggerated intros, they are mostly content-free fluff written for people who don't want to think. What you should do is dig out the information source, not rely heavily on those KOLs. Otherwise, in this market, who else would they harvest if not you?

Fourth, the value of systems and moats. Current tools like Fable are already powerful enough to let me easily cross industries; the only remaining barrier is computing cost. But the real moat cannot be crushed by sheer capital and technology alone—it requires time to accumulate. Just as TeraFab, no matter how strong, cannot surpass TSMC, because TSMC's strongest moat has never been technology, but "people."

If you haven't deeply used AI or experienced Fable yet (that subscription fee is hardly an investment), then in the future, you are the one waiting to be "allocated." Go use it now!

Next time, I might talk about some experiences with arbitrage and the pitfalls to watch out for.