Recently, while chatting with a friend in Silicon Valley, I was impressed when he told me, with almost incredulous tone, that the things AI can do now exceed your imagination. After hearing this, I casually asked my question: If that's the case, why does it seem like life hasn't changed much because of AI?

A friend explained that AI is still in the stage of rolling out infrastructure; on one hand, the valuation is higher, and on the other hand, mature infrastructure is needed to translate the changes brought by AI into practical applications.

Looking at the crypto field, Helium, which once made the market question its usefulness, has proven at the end of last year and the beginning of this year that Crypto is not just about tokens; the products themselves can also enter many people's daily lives. Like the long-tail market that cuts into communication, Crypto is also attempting to cater to small-scale training needs and AI training markets involving certain on-chain data through decentralized computing, storage, and token incentives.

Targeting Telegram, utilizing a decentralized computing network and cloud storage network to provide infrastructure for AI, focusing on the continuously hot meme in this market cycle, PinGo has carefully selected an easy-to-understand yet not easily thought of niche.

Construction of decentralized AI infrastructure

Currently, public information about PinGo is relatively limited. However, based on the project's documentation, we can roughly outline PinGo's vision for a decentralized AI infrastructure.

Aggregating idle CPU computing power

In terms of computing power, PinGo has established a network that can activate idle CPUs. Although this initially puzzled me, upon reflection, it indeed makes sense. There should be no idle GPUs in the current market; they are either rented for AI training or needed for specialized tasks such as rendering images, where 'idleness' is hard to come by. Only CPUs, being multi-threaded processors, have some computing power that can be considered 'idle.'

Although CPUs are not as computationally efficient as chips specifically designed for AI training, PinGo itself does not target overly complex models. As a supplement to centralized GPU computing rental, it can still meet the vast majority of needs as long as enough idle CPUs can be mobilized.

CDN-based "fog storage" and "fog computing"

In the documentation, PinGo mentioned plans to use CDN to quickly centralize CPU resources. CDN stands for Content Delivery Network, which is a network set up by traditional cloud services to improve data transmission efficiency by storing popular data in various locations, allowing users to avoid fetching data from the main data center each time, thereby alleviating pressure on the data center and greatly enhancing user experience.

Although PinGo did not explain in detail how its so-called CDN network operates, from an overall concept and architecture perspective, this CDN network is likely designed to enable quicker responses in a peer-to-peer network for training.

In simple terms, PinGo does not have a 'data center' for cloud services; it must have numerous 'data relay stations' to handle global storage and computation resources. These CDN nodes, upon receiving training tasks, will use algorithms to aggregate the nearest idle computing resources to process the data for training purposes. After training is completed, these CDNs will promptly feedback the results to the corresponding users.

Essentially, PinGo utilizes a distributed network architecture to meet the demands for data storage and computation during training, which is also the 'Decloud' concept it mentions. I prefer to describe the specific implementation process with the 'fog' concept, which is more decentralized compared to the cloud.

In terms of security, PinGo employs passive measures including NGFW firewalls and role-based access control, as well as active means like penetration testing and database auditing to ensure data security within the CDN network.

Low cost and customization

By leveraging idle computing power to build a decentralized data network to meet the demands of the AI long tail market, PinGo aims to achieve lower computing costs and more personalized customization needs through this solution.

Many people may wonder whether the demand for personalized customization can really not be satisfied given that a few major cloud service providers almost dominate the market. Is there really still market space?

This has actually been a question I have been puzzled about for a long time, but in fact, this market space is still larger than imagined. A report by Canalys showed that as of the fourth quarter of 2023, excluding Alibaba Cloud, Huawei Cloud, Tencent Cloud, Baidu Cloud, AWS, and Tianyi Cloud, the market still has a 4.4% share. Considering the nearly 10 billion USD spent on cloud services domestically in 2023, this share also amounts to several billion RMB.

I also consulted industry professionals about this, who indicated that large companies' cloud services have a low degree of customization. There might be a certain level of 'big fish bullying small fish,' as users can only choose services based on the tiers set by them. Additionally, the prices of these large cloud services are relatively high, which is partly because they can guarantee stability to a significant extent; however, many needs do not require such high stability. These ordinary needs have given rise to the existence of small cloud service providers.

For PinGo, even if its services only occupy a certain market share in the niche fields of Web3 and AI (with the international market size far exceeding 10 billion USD), it is sufficient to generate a very considerable cash flow that promotes on-chain activity. Its strategy of aggregating idle computing power not only lowers prices but also satisfies more granular needs due to the fragmentation of computing power, thus attracting user engagement. Just like Filecoin, which is widely used in data and information storage, ordinary users just lack sufficient perception.

Starting from memes, aiming for various AI tools

After the rough era of Web3 entrepreneurship, starting from a specific use case became a question that needed to be considered early on. PinGo seized two absolute hot spots in the current market: the TON ecosystem and memes as entry points, primarily focusing on providing Telegram users with AI-generated works early on.

Telegram bots and AI creation factories

The early accumulation of seed users requires a certain degree of incentives. Fortunately, TON-based projects have the huge user pool of Telegram; the project side only needs to think about how to find its target users among them. To this end, PinGo launched the Telegram bot PUNNY to help users participate in the early incentive program and the AI creation factory PinGo Gallery.

Currently, the plan to attract early users for PUNNY has been completed, with official website data showing it achieved a total of 500,000 users and 280,000 monthly active users, marking a very successful marketing campaign.

PinGo Gallery provides AI-based creative tools, allowing users to create based on keywords and share in Telegram. I believe this tool could even serve as a basis for creating Meme tokens. When certain popular memes emerge, users can quickly use the tool to create Meme images and issue Meme tokens based on that, spreading them directly in Telegram.

Whether it's simply creating memes in conversations or eventually pushing them to the market in token form, this tool has its uses and can be considered one of the early tools to attract users.

"Only contributors have the right to use"

In the design of PinGo, not everyone can use the decentralized computing network; only those who contribute computing power to the network and become participants have the right to use it. Participants can freely choose the amount of computing power they contribute, and after providing computing power to the network, they will receive an NFT as proof. Only by holding this proof can they use the computing power for AI training and utilize storage services for data and algorithm models.

This design largely avoids the network imbalance caused by many 'freeloaders' due to a lack of computing power contributors. In past experiences, such imbalances can lead to chronic network failure. While PinGo's mechanism design may prevent a certain kind of 'Sith attack' that leads to a short-term surge in users, it essentially plays a positive role in the long-term development of the project.

Ultimate goal

PinGo's goal is naturally not just to act as a small tool on Telegram; in the future, as the network scales, PinGo could provide developers with the capability to create smart contract proxies, AI investment advisors, AI sentiment trading tools, and other AI products based on information such as chat records on Telegram.

In addition to the product capabilities themselves, PinGo's roadmap indicates plans to develop a blockchain based on its CDN network next year and further build it into a second-layer network for TON. Developing an independent network will give the token PINGO and the project stronger value capture capabilities, and as the frequency of network usage increases, the demand for PINGO as the network currency will also continue to rise.

Once everything is completed, PinGo will become a large-scale AI infrastructure that can call upon computing and storage capabilities based on demand, providing low-cost resources for users who need training, allowing them to avoid purchasing expensive centralized computing and storage services. For a considerable period, the tokens can serve as an incentive resource for suppliers, which alleviates the financial pressure typically associated with traditional data centers, enabling rapid expansion.

Since we're talking about tokens, let me add that the PINGO team and investors hold less than 20% of the shares, with 80% of the shares allocated to incentivize computing power contributors, AI users, and CDN nodes, making the builders and users of the network the ones who truly decide the project's ultimate direction.

Decentralization is becoming an important supplement to technology infrastructure

It may be a pipe dream to say that decentralized infrastructure can replace large-scale data centers in a short time, but in reality, a number of DePIN projects have made the market aware of their important supplementary role. For certain not overly complex needs, services provided by Web3 projects have gradually been accepted by the mainstream market.

The reason why users in domestic markets do not have enough perception may be due to the maturity of infrastructure such as communication and payment. However, the fact is that in many capitalist countries, the level of infrastructure construction varies greatly across different regions of the same country due to a lack of adequate incentives. This is also why DePIN, stablecoin payments, and others being core areas of Web3 seem incredible to us.

With billions of people worldwide, the egalitarianism brought by Web3 can bring the latest technology to thousands of households and reach economic benefits to corners that capital may overlook. Perhaps in a certain corner of an island nation in the future, Web3 infrastructure can help people there connect to the internet and use computing power that they could not originally afford to fulfill their small yearning for technology.