In 2024, besides the Bitcoin ecosystem, the hottest track is DePIN. DePIN stands for Decentralised Physical Infrastructure Networks. Sun Jing has been leading the IoTeX team in this track since 2017, and Sunny, an investment partner at Hashkey, also paid early attention and investment to this track.
Part1: Evolution of the Concept of DePIN
Q1.Mempool: Please explain the concept of DePIN. There are various definitions of DePIN in the market. Some reports suggest that DePIN refers to replacing existing infrastructure through hardware investment and token incentives, while Messari believes that DePIN deploys real-world physical infrastructure and hardware networks through an encrypted protocol. Can you explain your understanding of the concept of DePIN and its core elements?
Sun Jing: Recently, DePIN has sparked widespread discussion. For those who are new to the crypto field, DePIN may be a new concept. DePIN is an abbreviation for Decentralised Physical Infrastructure Networks, which translates to "decentralized physical infrastructure network" in Chinese. The simplest explanation is that it is a physical infrastructure network created and maintained from the bottom up by the community. The core elements of DePIN include:
Decentralized: Unlike infrastructure networks led and deployed by centralized institutions such as governments, DePIN emphasizes the participation of communities from different regions and countries in infrastructure construction.
Physical infrastructure: Different types and sizes of hardware, such as IoT, 5G information towers, various servers, or mobile phones.
Token economy: A network incentivized by the economy to contribute and build together.
Regarding whether this definition emphasizes the on-chain data, we should note that DePIN includes not only data networks but also resource networks, information networks, computing power networks, etc. DePIN networks have diverse use cases, and data is just one of them.
Sunny: The term DePIN was widely adopted only last year, but exploration in this field had been ongoing for a long time before the term emerged. Broadly speaking, the concept of DePIN integrates two core elements:
Basic physical resources: including the most basic IoT and industrial IoT, as well as the storage, computation, and transmission of data. Recently, this concept has expanded to a broader range of resources, such as energy or infrastructure construction.
A highly complex yet dynamically balanced token economic system: this system effectively incentivizes interaction between resource providers and demanders.
Q2.Mempool: DePIN was proposed by Messari in November 2022. At that time, they conducted a poll on Twitter, hoping that people would choose a name for this track from several candidate words. DePIN stood out from a series of candidate words. Subsequently, the concept quickly gained widespread recognition and popularity in 2023. Will such a concept proposal, even with some changes in its connotation and extension, have an impact on entrepreneurs and investors?
Sun Jing: To understand this issue, let's first briefly review the history of IoTeX.
IoTeX can be said to be the earliest promoter of the DePIN track, and it is also the earliest public chain project in the DePIN track. Its original intention is very consistent with DePIN, but the concept of DePIN did not exist at that time. IoTeX's original intention was to create a decentralized IoT, emphasizing the connection of various devices and linking events, information, and assets in the physical world to the blockchain. At that time, the blockchain played the role of a decentralized ledger, and its most important function as an open financial infrastructure was not realized.
It was not until the second half of 2020, when the public chain matured and the infrastructure was ready, that DeFi began to emerge and lead the application development of the entire crypto industry. After DeFi, GameFi emerged. At this point, we realized the important value of blockchain as a value circulation system.
By the end of 2021, IoTeX proposed the concept of MachineFi, which was seen as an early form of DePIN. The biggest difference between MachineFi and the previous decentralized IoT is the introduction of a token economy, using the blockchain to realize value transfer, thereby unlocking the potential economic value of the IoT in the Web2 physical world.
Gradually, we found that different investors and practitioners also saw the same vision. Some proposed Token Incentivized Physical Network, or TIPIN, and some investors proposed EdgeFi, emphasizing the facilitation of a network through edge computing devices. Everyone proposed different words until Messari proposed the concept of DePIN, which gained industry consensus. Through this term, everyone with the same vision was united, linking and uniting various resources, gradually forming the DePIN track.
Sunny: From the perspective of an investment institution, HashKey Capital is committed to finding a field that can lead to mass adoption in society, such as token economics, blockchain technology, and application-layer possibilities such as gaming, SocialFi, and DeFi, all of which have the potential for large-scale adoption.
The logic that prompted us to invest in IoTeX's MachineFi was exactly this: assuming a simple IoT device such as a watch or a mobile phone, if each phone could join an application related to crypto, its adoption would actually grow to millions or tens of millions, far exceeding pure crypto applications. From this perspective, DePIN is a natural track for achieving large-scale adoption.
While focusing on this track, we have also seen a lot of progress and changes, such as projects like IoTeX, Helium, Random, which have already completed a very important wave of market education. Through this market education, we have seen in the past few years that in the fields of data storage, computation, and transmission, or in more practical applications such as car batteries, energy, and other different niche tracks, many new projects have emerged.
At the same time, we have also seen more people adopting this concept, including some practitioners in traditional industries, such as the telecommunications and energy sectors, who are willing to try to introduce token economics into their projects and test whether this token economy can generate a flywheel effect and activate the entire industry's resources. In turn, these entrepreneurs are driving capital into this track. Therefore, we see that a positive cycle and interactive relationship is forming between capital and business.
Q3.Mempool: Everyone mentioned the outbreak in 2023. Why was it 2023? Did infrastructure or other aspects reach a turning point?
Sun Jing: I remember the concept was proposed in November 2022, and in the early stages, there were small-scale discussions among the builder community involved in the development and construction of DePIN projects, but the concept did not become a track, and there was little attention from capital.
At that time, IoTeX wanted to connect everyone and organize an event to allow the market to see this new track and understand its differences from other tracks such as DeFi and GameFi. IoTeX initiated an event called "R3al World" at EthDenver in 2023, bringing together all DePIN projects in the market and providing them with a platform to showcase their products. This event was very successful and was the first large-scale event in the industry focused on the DePIN track. Many participants were exposed to this track for the first time, and one review was quite interesting, calling it "the WEB3 version of CES" (CES is the largest consumer hardware exhibition in the United States, held in Las Vegas every January).
Although the scale of the event was small at the time, with only two or three dozen projects, the track had already begun to take shape. This was mainly due to developments in two areas: first, blockchain technology had matured, and DePIN projects could be verified on-chain; second, the infrastructure for DeFi had also been established, and overall, each project began to seek and position its own use cases, gradually advancing development. However, at that time, there was still a lack of infrastructure to transfer data from off-chain to on-chain, which was the key infrastructure—W3bstream—later led and launched by IoTeX.
This event was just the beginning. Subsequently, through continuous industry communication and events, investors began to see its potential and invested more funds in the market, and more entrepreneurs entered this field. At the end of last year, we launched the DePINscan data platform and were surprised to find that the number of DePIN track projects in the market had grown from the initial two or three dozen to over 600.
Sunny: I am curious about where the entire on-chain interaction is concentrated? Last year, especially in the second half of the year, the trading volume of the entire DePIN track continued to grow. For example, Arweave's storage volume reportedly had 1.8 billion retrievals, with 1.2 billion of them occurring last year. But in my opinion, most of the interactions actually occurred due to pure Crypto projects. From your data, are there more Web2 IoT projects or cross-circle projects paying attention to Web3 solutions, and do you have more insights?
Sun Jing: Currently, the data is basically on-chain interaction, not including off-chain. Because of the many projects that have grown this year, the gradual maturity of the projects will prompt the network to start, which means that token incentives will begin to generate, and on-chain interaction will also occur.
This is also why we believe that the DePIN track is conducive to large-scale adoption. First, it incentivizes users to contribute devices through continuous token usage, and each token incentive will form an on-chain interaction. Second, it truly implements an economic model. Unlike many on-chain projects that only have an economic model in theory and do not execute it, DePIN projects have a fully blockchain-controlled economic model on the supply side, which can be likened to the operation of Bitcoin, clearly defining when and how the mining mechanism is triggered.
We had hoped that Web2 companies could enter this track and recognize the Web3 economic model, but reality has proven that we were too optimistic. Web2 and Web3 are basically two parallel operating modes. However, Web3 companies still have the opportunity to collaborate with Web2 companies—especially those with a large number of IoT devices. DePIN can directly utilize these devices. In the Web2 economic system, these devices are usually controlled by centralized institutions for data and usage rights, but in the world of DePIN, each device purchaser is not only the owner and user of the device but also a contributor to the network, and can earn income from their contributions, forming a continuous process.
Therefore, we believe that for quite some time, the innovation in DePIN will mainly be driven by companies rooted in Web3. At the same time, the founders of these companies often have deep experience in the IoT or Web2 industries and a profound understanding of the Web3 economic model, enabling them to leverage unique advantages in this cross-domain.
Sunny: From the conceptual level, applications, and on-chain activity, we can see a strong surge in DePIN in 2023. From the overall macro environment, we expect a stronger surge in the DePIN track in the next two years. There are two reasons for this:
First, in terms of the qualifications of entrepreneurs, it is only possible to succeed if there are people with Web2 industry experience and resources who also understand Web3.
Second, from the market environment, the SEC approved an ETF in January, which is equivalent to opening a door at the regulatory level. Not only can investment institutions enter the market smoothly through ETFs, but it also signals a more relaxed and friendly regulatory environment, meaning that many large entities interested in exploring the DePIN field can explore with more confidence. If the prediction is correct, the effect of this large-scale application can transcend bull and bear cycles because it has a solid source of income, clear supply and demand, and a clear token economy. We also hold a relatively optimistic attitude towards the surge of the entire track in the next two years.
Part2: IoTeX's Infrastructure and Economic Model
Q1.Mempool: Next, I would like Jing to introduce us to IoTeX's infrastructure, which I believe can help the audience better understand DePIN.
Sun Jing: IoTeX is probably the most advanced and largest infrastructure platform specifically designed for the DePIN track in the industry. This infrastructure platform consists of two main parts:
First, IoTeX's public chain, Layer1, is a key financial infrastructure in DePIN. In the DePIN system, it is crucial to incentivize various contributors in the network, including the supply side and the demand side, using a token incentive mechanism. All participants ultimately need to transact and settle efficiently on Layer1. This aligns with IoTeX's design philosophy of pursuing a decentralized IoT—having high TPS (transactions per second) while being able to connect hundreds of millions or even billions of devices in a compatible manner, enabling these devices to settle on-chain simultaneously.
Second, in 2023, we focused on off-chain verification and computing facilities, as middleware or Layer2, to bring physical world data onto the chain in a verifiable manner. However, not all data needs to be stored on Layer1. In most cases, only the proof of data verification needs to be brought on-chain, while computation and storage are done off-chain. This involves a key requirement in the DePIN network: zero-knowledge proofs (ZKP). For example, a user's vehicle data, such as information about when the vehicle arrived at a certain location, is crucial for demanders such as insurance companies, who use this information to set insurance rates. However, users do not want all personal information to be disclosed on-chain, so ZKP technology is needed in this case.
In addition to the above, IoTeX is currently deeply involved in two things:
First, how to enable DePIN project parties to develop projects better. This involves many development tools, which can greatly shorten the time from development to launch for DePIN project parties, from two to three years to two to three months or even weeks.
Second, how to better serve DePIN users, including device users and miners, making it easier for them to participate in different projects.
Overall, IoTeX provides a modular infrastructure platform specifically designed for DePIN.
Q2.Mempool: IoTeX was founded in 2017, focusing on the decentralized IoT field at the time. Before that, although the concept of IoT was widely discussed, the progress of actual applications was relatively slow. As an entrepreneur, did you really feel the accelerating effect of crypto technology on the development of the IoT industry? Please select a few important milestones from IoTeX's development history to discuss how crypto technology has helped the development of the IoT industry.
Sun Jing: Before doing Crypto, I was in venture capital, mainly focusing on AI, IoT, VR, AR, and other fields, covering many innovative tracks. However, in the process of investment, we found that regardless of the type of startup company, there would be bottlenecks after reaching a certain stage. There are two reasons for this:
First is the bottleneck of startup companies. For example, all AI companies face challenges in data acquisition—most of the data is controlled by centralized enterprises, and innovation in AI models alone cannot scale these companies into a viable industry. Similarly, IoT companies require significant initial capital to establish a useful network. They need to invest a large amount of capital to deploy physical devices and bear high maintenance costs, which is why in the early stages, only large-scale companies could enter the IoT field, but with low efficiency.
Second is the bottleneck of centralized companies. In the traditional centralized model, the value of IoT data is subject to many limitations and cannot be fully realized. This data is controlled by centralized companies, making it difficult to introduce more partners and service providers into the ecosystem through open-source or more open means, thus setting the growth flywheel in motion.
The programmability and open-source nature of encryption technology, as well as smart contract systems, have shown us new ways to build IoT systems in the future. The vision of IoTeX at its inception was to build a decentralized physical system. Although our original intention was very ambitious, the actual process of promotion started from scratch and went through several key development milestones:
Launch of the public chain. In 2017, the infrastructure of blockchain was still very weak. Although Ethereum had already been launched, its functionality, performance, and speed were still unable to meet the needs of IoT systems. Therefore, we spent three years building IoTeX's public chain, creating a dedicated public chain that could meet the needs of IoT.
Deployment of on-chain devices. After the launch of the public chain in 2020, we did not only view blockchain as a financial infrastructure but also emphasized its decentralization, security, and privacy. In that year, we launched the industry's earliest on-chain hardware device, Ucam—a smart camera completely managed by blockchain identity. This product received positive feedback from the market and is still in circulation. Although this step has not yet fully integrated into the economic model, the key is the integration of hardware, hardware ID, and middleware.
Introduction of the MachineFi concept. By the end of 2021, we realized the importance of financial incentive mechanisms in centralized IoT. Therefore, we introduced the MachineFi concept, largely benefiting from the maturity of DeFi and other NFT scenes, which have become financial modules serving other tracks. At that time, we saw the scene of the DePIN track: unlocking the economic and financial potential in decentralized IoT, realizing the financialization and value circulation of physical world devices, resources, and data on the chain.
Development and strategic positioning of DePIN. By 2023, after Messari proposed the DePIN concept, the industry quickly formed a consensus. At the beginning of the year, we made DePIN first an important strategy for the entire network. Currently, we have reached cooperation with about 70-80% of the projects in the industry, becoming the largest and earliest modular infrastructure platform in the industry. We have launched a series of modules, including off-chain computing W3bStream, information data platform DePINscan, device identity management Identity, and established a Liquidity Hub to better incentivize users to use products and trade assets, providing diversified services and modules for the entire industry.
Q3.Mempool: From the perspective of implementation, has IoTeX brought about new usage scenarios that have changed users' lives? Or is it still mainly focused on infrastructure?
Sun Jing: IoTeX mainly supports various types of innovation in the DePIN field through infrastructure, with a special focus on AI-related DePIN scenarios. We believe that AI-related DePIN may become the fastest-growing and largest-scale sub-track. For example, there is a project called BiTTensor, which has built a whole set of technology stacks based on open-source AI models and opened them to various AI developers to create various subnets that can provide diverse services. In addition, we have also seen many AI-based computing platforms and data platforms that handle a wide variety of data types, including geographic data, biological data, and retail data serving various scenarios. The core of these applications is still closely related to the collection and use of decentralized AI data, which we believe has enormous potential. We also pay attention to vehicle data. Many companies are focused on the mobility sector, each with its own focus. Some DePIN projects focus on decentralized surface, while others allow users to contribute vehicle data to create decentralized high-precision maps, similar to a decentralized version of Waze, and some are working on decentralized high-precision maps, and so on.
In the future, IoTeX may launch some more daring experimental platforms to bring more devices onto the chain. For example, the recent popular Mingwen is actually creating a new fair distribution mechanism that allows users to participate, but this distribution mechanism is easily disrupted by large studios. If there is a layer of authentication based on physical addresses and device identities, the distribution mechanism of Mingwen will be fairer.
We have also seen some companies in the energy sector attempting decentralized energy. For example, there is a company called PowerPort, which installs a device for electric cars and then obtains data on your use of green energy through the device. This data is itself a credit in developed countries like the United States, and this company incentivizes users based on the time periods when they use green energy, encouraging users to charge during periods of low electricity consumption.
There is also a track involving the use of existing wearable devices, such as watches and phones, as data collection tools, also known as BYOD (bring your own device). These devices act as terminals for collecting user's daily data, which is then contributed back to the user, forming a closed-loop network. However, this field also faces some challenges, such as how to ensure that users upload valuable data and how to effectively utilize this data. But we have also seen some good scenarios, such as a company on our platform called WifiMap, which has established a network that allows everyone to share their own Wi-Fi resources and use other Wi-Fi resources on the market. The network currently has 10 million real users.
Q4.Mempool: The economic model does play a very important role in this track, as the existence of the economic model enables the formation of this decentralized network. What are the better models in the industry, and what are the common elements?
Sun Jing: Tokens play a crucial role in the entire decentralized network, not only accelerating the circulation of value within the network but also quantifying the value of different contributors, service providers, and users in the network. IoTeX itself is a public chain, which, in addition to the basic payment network Gas model, also has a staking model.
Firstly, there is the Gas fee charged for each transaction, which is the fee required to use the blockchain network.
Secondly, there is staking to obtain network services. Staking first allows the network to reach consensus to enhance consensus security. In addition, staking gives stakers the right to participate in the network. For example, demanders can obtain standardized services provided by the network by staking the tokens of the DePIN network, such as verification or data services. For example, IoTeX's W3bStream is an infrastructure service that provides off-chain decentralized computing to verification. If different node operators want to participate in W3bStream and provide services, they can do so by staking IoTeX tokens; for project parties that need these services, they also need to stake IoTeX tokens to obtain these services.
In addition to this economic model in IoTeX, there is also a Burn-and-Mint (BME) model for obtaining services, which converts tokens into a unitized purchasing power. In the DePIN track, Helium is one of the earliest projects actively promoting this BME model. In this model, the network has two types of tokens: one is the token of the network itself, and the other is a general-purpose token used within the network, which can be seen as points in Web2 companies or stablecoins used within the network. The reason why this stablecoin-like token is important is that the price of the network's token usually fluctuates greatly, but the purchasing power of goods and services is relatively stable. Therefore, users can burn a certain amount of tokens to convert them into a corresponding amount of stablecoins to purchase goods. Of course, there are many other application methods, but staking and burning are the two main ways, and their purpose is to enable participants in the network to participate in the network economy using tokens and provide a relatively rational and calculable pricing model for the external economic vitality of the network.
Part 3: Market Size and Integration of DePIN with the AI Track
Q1.Mempool: Messari's report suggests that DePIN's flywheel will significantly increase global GDP in the future, possibly reaching trillions of dollars in revenue. How do you view this market forecast?
Sunny: In my opinion, many areas are still in the exploration stage. For example, projects that have been dedicated to this field for a long time, such as IoTeX, Helium, and Render, have seen a continuous increase in data volume over the past few years. This is because large enterprises are gradually starting to shift some storage and computing capabilities to decentralized networks. At the same time, some car manufacturers and different device manufacturers are also beginning to integrate token economies into their devices. Additionally, we have seen native crypto projects, such as the launch of the Saga phone in the Solana ecosystem and Helium's efforts in promoting 5G, embedding applications into devices through the sale of phones, which are explorations on different levels. The essence of DePIN is supply and demand, which means a product must have strong market demand to attract corresponding purchasing power. Only when purchasing power continues to exist can the flywheel effect or economic benefits continue to strengthen and expand. Currently, many projects are focusing on the supply side, promoting the flywheel effect to stimulate mutual competition of supply-side resources, reduce costs, and attract market demand. However, market demand is a variable dimension involving user experience, cost, and other factors, so we are both optimistic and cautious about this track.
Sun Jing: We agree with Messari's prediction of future growth, mainly based on two factors:
Firstly, from the historical development of society, the proportion of machines and devices in global GDP has been on the rise. With the development of technologies such as artificial intelligence, the demand for human labor is gradually decreasing. We expect this proportion to exceed 40% in the future, forming a massive trillion-dollar market. However, the productivity formed by these machines does not yet have a unified way of value circulation, leading to the locking of data value, resource value, and network value. DePIN can largely unlock these previously constrained values.
Secondly, we need to consider the scale of the physical infrastructure market itself. We believe that DePIN represents a new type of community-driven, bottom-up elastic model, building infrastructure in a shared economy manner. Although many countries already have centralized infrastructure, decentralized infrastructure owned by the community is still needed as an alternative. For example, the current situation of war in Ukraine shows that centralized infrastructure is easily destroyed. If there is a set of decentralized alternative infrastructure that is difficult to control directly, it is extremely important for the survival of the country and even people's livelihood. Similarly, in the field of AI, the most powerful AI infrastructure is currently controlled by tech giants like Microsoft and Google, who use a large amount of resources to build centralized AI computing systems. If there is a decentralized system, whether it's models, computing power, or basic chips, it would be a form of protection for most ordinary people.
Therefore, although the DePIN sub-market may not directly compete with the most advanced centralized infrastructure in the short term, its market share will gradually increase. Even if it only accounts for 10% of all infrastructure, it will still form a market worth trillions of dollars.
Q2.Mempool: We see that on one hand, DePIN can bring solutions to AI; on the other hand, AI can indeed bring new narratives and demands to the DePIN track. What are your observations on this vertical field, and what opportunities might arise in the future?
Sunny: First, let's explore the applications of AI+DePIN: Since the launch of ChatGPT in 2023, it has brought revolutionary advancements to the entire encryption technology industry. We can analyze it from the following aspects:
AI technology has optimized the entire technology development process, making the technology stack more efficient.
DePIN, as an automated and self-maintaining IoT network, has enhanced the network's automation and decision-making capabilities by introducing AI algorithms and large-scale models. It has also strengthened intelligent identification of abnormal behavior, significantly improving the efficiency, automation, and security of the physical network of DePIN.
Next, let's look at the role of DePIN+AI in driving the development of the AI field, mainly from the demand of the AI technology stack:
Computing power. AI infrastructure requires powerful computing support, and currently, these resources are mainly concentrated in the hands of a few large companies. The key current need is to integrate these dispersed computing resources through a decentralized network, enabling small and medium-sized enterprises to access high-quality computing resources.
Data collection and storage. These can naturally be integrated into the physical network of DePIN, providing AI with more decentralized and even lightweight device-side storage and computing capabilities, in line with the original design of DePIN.
Models. Effective data models require data circulation and distributed computing capabilities, which is an area where DePIN can naturally meet the needs of AI.
Finally, we are also focusing on the popularization and practicality at the application level:
AI can make hardware more intelligent, making human-machine interaction more attractive and interactive.
Exploring how devices, algorithms, and token economies can be combined, these are natural integration points for AI and DePIN.
Sun Jing: AI is one of the most powerful productivity tools in human society to date, and DePIN represents a new way of infrastructure construction and economic distribution model, complementing each other. AI can be widely applied to the infrastructure of DePIN, making the DePIN network more intelligent, user-friendly, and valuable, thereby stimulating stronger demand.
Firstly, DePIN can serve AI in terms of computing power, models, data, and more, releasing capabilities that AI originally did not possess in a decentralized manner. For example, DePIN allows more people to participate, putting idle resources into AI network computing services. We have seen many DePIN AI computing platforms that enable people to use their idle resources for AI network computing services. In the decentralized AI field, although the current focus is on developing large general AI models, many practical needs are actually for smaller customized models, which require precise, real-time data. The DePIN approach can introduce the necessary data and address the data needs of these small models.
Furthermore, DePIN makes AI more open, transparent, and secure. For the general public and even AI practitioners, security is both a Pandora's box and a black box. Closed-source leads to opaqueness of problems and risks, and people can only rely on the opinions of top scientists and experts in centralized companies. But if AI becomes more open-source, then everyone can better understand internal issues, more effectively monitor potential risks, and collectively find solutions.
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。
