At the end of a previous article, a reader left a comment mentioning a question:

Is there an opportunity in the decentralized AI storage sector?

In the reader's comment, he mentioned the context behind this question, where things that it should remember are not retained, and things that it shouldn't remember keep reappearing.

From the perspective of the issues reflected by this question, I think it is strictly speaking a defect in the current AI technology related to contextual processing.

This defect is more related to the current AI algorithms, which are essentially not a storage issue—limited storage capacity or difficulties in scaling storage—but rather that the existing AI algorithms cannot efficiently store longer contexts and more precise, valuable contexts.

If we follow the current algorithms, to achieve sufficiently rich and accurate storage of contexts would lead to costs that the algorithm cannot bear.

Although the essence of the background issue reflected here is not a bottleneck in storage, it reminds me of a track in the crypto ecosystem that was once highly concerned but has "vanished" in recent years: decentralized storage.

Regarding decentralized storage, I still believe it is a significant market with strong demand. However, the reality is quite stark; this market has not been activated on a large scale to date, let alone developed.

If you participated in the crypto ecosystem in its early days, you surely remember Filecoin vividly.

The participation wave sparked by Filecoin could be described as a moment when the crypto ecosystem had the highest expectations for decentralized storage, pushing this track to its peak.

Although its final outcome disappointed almost all participants, I still believe that the vision and expectations it represented are very valuable and significant—breaking the monopoly of traditional centralized giants over storage and fully utilizing the scattered storage capacities around the globe.

This vision and expectation are reasonable from the perspective of market demand and business logic, in my opinion.

Many people have analyzed the outcome of this experiment in various ways, and I think there might be two reasons:

One is a technical reason;

The second is that perhaps the true breakout point of this demand has yet to arrive.

Let's first look at the technical aspect.

Blockchain technology essentially achieves reliable security and permissionless access through inefficient storage (every node has the same ledger records).

However, in various past and present decentralized storage projects, the part that truly utilizes blockchain technology is only in the accounting aspect, using it to distribute tokens.

The storage networks in these solutions (a network connected by peer-to-peer storage servers) do not actually use blockchain technology for the allocation and scheduling of storage capacity (even though their algorithms are also called "consensus algorithms"), and are more akin to traditional centralized distributed technology.

So, in my view, these solutions are just traditional centralized distributed technology with a layer of token incentive added.

Disregarding whether the technology used qualifies as blockchain, if this solution could truly activate or even realize this market demand, that would be quite remarkable.

But based on the current results, it has not only failed to achieve market effects but has also deviated from the original intention in its technical implementation—failing to aggregate the idle storage scattered around the globe, instead turning into competition among storage operation oligarchs.

So I have wondered:

Is it possible that the technology truly suitable for this sector, which is capable of distributing scheduling and incentivizing idle storage, is not blockchain technology?

Or is blockchain technology suitable for this sector, but the issue is that the various solutions provided so far have "consensus algorithms" with some insurmountable obstacles that may take several years to resolve?

Furthermore, looking at the breakout point of market demand, I have questioned whether this sector might follow the development of internet e-commerce, where the early attempts (like 8848) failed because the infrastructure and technical solutions at that time were not sufficiently mature and the market was also not mature enough? It wasn't until later that Alibaba appeared, hitting the right timing, location, and people, successfully igniting this sector?

In fact, it is not only decentralized storage; there are other decentralized fields (like computing power, data, etc.) that I see as having a large market and strong demand. Moreover, it is worth noting that some heavyweight developers in the AI field have also realized the danger of AI monopoly and proposed that decentralization is the only solution.

In reality, these ideas have been discussed and attempted in the crypto ecosystem for a long time. Unfortunately, these attempts currently have no typical success cases, and there are no good answers at present.

Perhaps only time and patience can bring us surprises at some point in the future.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。