The most outrageous claim in Web3 self-promotion is that "decentralized storage problems have been solved."

Take a look at the current situation—there's basically none.

Blockchain speed is competing, concepts are competing, but when it comes to handling large-scale data, the whole story collapses: on-chain costs are ridiculously high, off-chain trust becomes a black box, and returning to centralization is even worse. Everyone is aware of this awkwardness, but no one wants to discuss it properly.

Recently, a class of data layer projects has emerged attempting to directly tackle this difficult problem. For example, using erasure coding at the underlying mechanism level, which ensures decentralization and resistance to censorship while keeping storage costs within acceptable levels for practical applications. Once storage costs are no longer a bottleneck, Web3 will have the opportunity to truly reach ordinary users.

On the application layer, you can only see superficial prosperity; only by looking at the underlying infrastructure can you see where the ceiling is. These are two different levels of understanding—the former is today's hot topic, the latter is the potential for the future.