The size of a market and the quality of a business are related, but they are not the same variable. A market can expand rapidly while competition reduces the profit available on each sale.
That is the part of the memory story I do not want to wave away with an AI demand forecast. Demand can be enormous and still attract enough capacity, substitution or new competitors to change pricing. A supplier benefits from scarcity only as long as the scarcity survives the investment it encourages.
Reports of a technological advance need verification. A lab milestone is not automatically commercially competitive production. Yields, reliability, qualification and volume matter. But dismissing the report before checking those factors is just the bullish version of overreacting to it.
I would ask how much of the valuation depends on current margins persisting. If the thesis requires both sustained volume growth and a scarcity premium that never attracts an effective response, it has a demanding hidden assumption. The useful question is not whether AI is real. It is how long the supplier can keep earning exceptional returns while everyone else tries to participate.