Shi Jing ping

Una delle ragioni per cui Walrus è così significativo è che la criptovaluta continua a generare più informazioni di quante le attuali blockchain siano in grado di gestire, quindi abbiamo bisogno di un archivio o soluzione di archiviazione che ci permetta di conservare in sicurezza tutte queste informazioni per gli operatori. Gli operatori vedono la componente di archiviazione della tecnologia come un'"opportunità infrastrutturale"; tuttavia, Walrus è molto di più di una semplice rete di archiviazione generica. È una piattaforma sviluppata per gestire grandi quantità di informazioni, inclusi file, multimedia, dataset di intelligenza artificiale e persino lo stato delle applicazioni. Walrus è inoltre progettato per consentire ai tipi di dati menzionati di essere verificati, prezzati e utilizzati all'interno di applicazioni OnChain, invece di semplicemente archiviarli "da qualche parte" o perderli per sempre. Le scelte che abbiamo fatto durante la progettazione di Walrus potrebbero sembrare più incentrate sul piano tecnico; tuttavia, hanno un impatto significativo su quanto l'intero ecosistema potrà crescere senza costringere ogni singola blockchain a diventare un disco rigido estremamente costoso.

My understanding of the importance of "Walrus" didn't come through a chat about its price or some long thread about the market, but rather was born out of a simple yet universal challenge faced by anyone building an AI application or creating their own NFT: How do you protect the data (files/images, etc.) in the event that someone takes control of them all of a sudden and no longer allows you access to your files? As simple as it might sound, Walrus is already being used in the AI and NFT space simultaneously for just this reason. Walrus is a decentralized way to store large unstructured files (such as audio, video, PDFs, images, etc.) and is using Sui as a co-ordination layer that is used to manage how their storage network operates. The intention of the project is to provide reliable, valuable and governable access to the data in the world, as opposed to simply having the data hosted somewhere.
Both traders and investors are asking the same question about the sudden importance of storage. The reason is that both AI & NFTs are very data-heavy and can be easily damaged if not stored on-chain or behind a centralized entity. With NFTs, we typically think of the "NFT" being the artwork itself; however, in many cases, the actual artwork is stored off-site and referenced via a unique id in the NFT. Therefore, if those files or storage were to disappear or change, it would effectively render the NFT useless. In the case of AI, data sets, model files, logs, and agent memory are also all very large files, and there aren't realistic ways to store them on a standard blockchain. Walrus proposes to provide a layer where such data can be stored, while still enabling the verification of what was stored on-chain and when it was stored on-chain.

The Walrus project embodies an additional aspect of current AI evolution; the establishment of data markets. As AI making models and agents reliant on continually acquiring new datasets, those that prove Quality, Provenance, and Availability of data will become critical to AIs. Walrus has positioned itself as a foundational platform for monetizing and verifying data, as well as following the current trend of utilizing Auditable vs "Just Trust Me" types of data inputs into AIs.
In addition, while the NFT community has clearer immediate interest in a functioning NFT Ecosystem because of the need for fast access to Media and Metadata at scale, it will also require Walrus to fulfil storage needs for a range of products to include Stickers, GIFs, In-Game Assets, and other evolving forms of NFT collections. There's a long list of notable NFT businesses utilising Walrus for the storage of large Digital Content Libraries, on the basis that these collections are designed to contain such a high volume of heavy media; i.e. if you've ever attempted to load an NFT collection whose half photos timed out or whose metadata has become corrupted, you know how fast trust erodes.
A good real-world example that illustrates the point without the "marketing" spin is that if you have a small indie game company with 10,000 character NFTs. Each character has many different skins, animations, audio, etc. With time, the expectation of players is to have fresh content regularly and with seasonal updates. The players may also expect more from their experience, for instance, AI-generated alternate versions of the characters.
The company could host the media file on a standard cloud service, however, that brings an additional layer of risk because if hosting costs increase, if the company gets bought out, or if cloud policies change, the NFT assets could be limited, deleted, or replaced.
By using an infrastructure solution such as Walrus, the company transforms its "trust us to always monitor and pay for hosting fees" to "our assets are residing on a network built to ensure permanence and verifiability." Storing files via this method does not eliminate every risk; but it changes the failure mode; therefore, this is an important aspect that should be taken into account by each investor.
The second half of the story relates to Artificial Intelligence (AI). This is not just a theory. Partnerships have chosen Walrus to be the storage layer for partnerships that require their agents to utilize large binary files within their workflows. In other words, the agents cannot be effective unless they fully take advantage of all the memory, data, and artifacts they have access to.

The continued discussion of Walrus in relation to NFTs and AI is primarily due to the fact that Walrus is developing an economic framework with per-use pricing, incentives, and community alignment as the basis of its system and that the WAL token is being deliberately distributed across multiple mechanisms (60%+ through airdrops, subsidies, and reserves meant for community use) with a focus on distributing WAL to individuals who will be active members of the community. This method of token distribution may not necessarily guarantee success; however, it will directly impact how members of the ecosystem participate within the Walrus community and may lure more builders into creating applications and tools that support Walrus rather than simply investing. Additionally, the airdrop mechanism associated with the soulbound NFT functioned as the initial introduction for Walrus into NFT culture. Regardless of people's opinion on airdropping, employing an NFT as a qualification for receiving WAL provides an example of how membership-driven engagement in an NFT environment creates closer ties between the participants; thus, small factors such as this contribute to the form and development of communities, which will ultimately have a significant impact on the ability for networks to successfully achieve initial network effects.
The scale metrics Walrus has released publicly since launching its mainnet in March 2025 detail that it stores hundreds of terabytes of data with millions of blobs. Investors view adopted signals like these as more important when assessing the value of storage networks than as a means to create earnings on a quarterly basis. Storage networks become considered an asset when a user saves critical information using them repeatedly.
Some risks of decentralized storage have been noted. Decentralized storage is a business area that is very competitive. User loyalty is not guaranteed for users of decentralized storage. Increases in costs, unacceptable performance levels, or difficulties associated with building developer toolkits that work well together could result in builders choosing to find alternatives. In addition to the risks associated with any competitive industry, there is a concentration risk due to heavy reliance on the ecosystem as a whole. Walrus is part of the Sui stack, which can be a strength for Walrus but also is a significant limitation. How quickly the Sui stack expands can impact the future growth of Walrus. An additional point to consider in determining the future demand for decentralized storage as compared to the level of growth in the AI market is whether companies will still want to store the most sensitive information on-site or utilize blockchains for selected types of information.
The Walrus project has gained positive attention in both the AI and NFT communities because it solves an issue these communities are facing in common. Using the Web3 builder’s perspective on ownership, permanence, and verification, Walrus offers a solution that is true to the Web3 community. If you’re considering Walrus as a trader/investor, rather than viewing it as a meme or a pure infrastructure investment, try to think of it as a wager on the future of AI agents and digital media economies. If these types of economies move toward a storage layer for their data that is not merely stored, but actually verified by third parties, then the storage layer will not just be a background element anymore; it will become one of the key factors on which to place your trust.
@Walrus 🦭/acc #walrus $WAL