The U.S. stock market actually operates on a fairly "rough but effective" logic: buy whatever is missing.

When the market lacks memory, funds flow into memory manufacturing companies like Micron Technology.
When chips are scarce, semiconductor stocks become the focus, from design to manufacturing, all are highly sought after.
Signs of a bubble are not too hard to recognize: when supply begins to surpass demand.
But currently, the story has not reached that stage.
Technology companies are still rapidly expanding factories and increasing capacity.
The reason is very clear: demand is far exceeding supply.
The Artificial Intelligence wave is creating an unprecedented resource shortage.
A series of startups, teams, and AI projects are continuously emerging, all needing GPUs, chips, and computing infrastructure.
This causes the hardware supply chain to remain under constant tension.
However, the long-term outlook is not as "rosy" as it is now.
Among the countless AI projects flourishing, only a few (perhaps 2–3 names) are capable of surviving and generating sustainable profits.
Most others will be phased out.
When this cleansing wave occurs, hardware demand will drop sharply.
Factories built to meet the AI boom will face overcapacity, forcing production cuts.
But this scenario is unlikely to happen immediately.
It may not occur until around 2027 that the market truly enters a "cooling down" phase.
And at the present time?
The trend remains very clear: high demand – insufficient supply – strong growth.