Author: Shenchao TechFlow
As BTC breaks through $100,000, more funds are looking for new projects and opportunities under bull market expectations.
But if you ask which track currently has the most opportunities? AI Agent must have a name. However, with many AI Agents being launched every day, the narrative of the entire track is also gradually stratifying:
One type revolves around AI Agent applications, with corresponding tokens representing memes or the purposes of that Agent; the other focuses on providing the infrastructure that enables AI Agents to perform better.
The former has become crowded and competitive as it is easier to observe at the application layer; while the latter has relatively more space for breakthroughs.
What other capabilities do AI Agents urgently need?
Perhaps we can find answers from the recently popular 'AI KOL' aixbt:
Research has found that what aixbt says is not always correct; it cannot distinguish between true and false, cannot require experts to verify its assumptions, and cannot question itself.
Essentially, because aixbt is actually a large language model, it can only scrape and summarize from various public data, making it more like a repeater of aggregated public information.
So, if you can provide more diverse, personalized, and private data to these AI agents, it might perform better.
For example, sharing insights about trading low-market-cap altcoins, or investment strategies only discussed in paid groups for it to learn... But this data is not on the table; the aixbts cannot access it.
Note that this world does not lack enough data, but high-quality data is hard to obtain.
In the current craze for AI agents, the infrastructure for data is actually lacking.
Here, a narrative space and information gap is that if a project can collect more personalized and individualized data and feed it to the AI agents or organizations in need, it may find a unique ecological niche in this wave of hotspots.
Two months ago, we wrote about a project called Vana, which collects various types of data not available in the public market using a DAO approach, while also incentivizing data contributions and guiding the purchase and use of this data through tokenization.
It was just that the AI Agents at that time were not as popular, and the usage scenarios of the projects seemed less defined. In this wave of AI Agent hype, Vana has clearly found more opportunities and a more coherent environment.
Coincidentally, Vana is about to launch its mainnet and release its own token $VANA, and Vana has also updated its white paper and token economics, providing more detailed explanations of current data issues and its own positioning.
In the crypto market, timing is crucial. What new dynamics and changes in Vana are worth paying attention to now? Do tokens have more positive expectations?
We read the newly released white paper and will quickly help you understand the current Vana.
The data 'double spending' problem, blind spots in seeking benefits
Undoubtedly, everyone is chasing the profits in the AI agent craze.
Anyone can easily create AI agents, and the assets corresponding to AI agents can also be easily tokenized... But apart from purchasing the tokens corresponding to the AI agents, what other benefits can you derive?
This issue signifies new opportunities for individuals and new narrative spaces for projects.
Don't forget that AI agents may be using the data you contributed to train themselves, but you haven't earned a penny from it. For example, the aforementioned aixbt analyzes crypto hotspots, one source of which may be an article you wrote on your Twitter.
Therefore, opening Vana's new white paper, one concept in the first few pages quickly caught my attention: the 'double spending' dilemma of data.
Double spending, does this term sound familiar yet strange?
This concept actually stems from the double spending problem (double spending) solved by Bitcoin—preventing the same Bitcoin from being paid twice.
Bitcoin's solution to this problem is to record the details of a transaction on a public blockchain, which acts as an immutable ledger. Everyone knows the entire historical flow of a coin, ensuring that a coin can only be spent once in its current state.
However, in the data field, this issue is more complex.
Unlike Bitcoin, data is inherently replicable, leading to an economic dilemma overlooked in the AI craze: when data is sold directly, buyers can easily copy and redistribute it, causing the same data to be utilized multiple times, and you cannot gain any additional benefits from this utilization.
For example, if you wrote a tweet, once it is used and learned by an AI agent, it may be shared indefinitely with other AI agents, ultimately leading to the data losing its scarcity and economic value.
If you want to create a ledger like Bitcoin to record data usage on-chain, to avoid the double spending problem, is that feasible?
First, data itself can sometimes be private, making public records inappropriate, and you may not want to share; second, even if you record data usage, you still cannot guarantee that this data continues to be copied and resold off-chain. Third, everyone wants to take advantage of your data; who would want to join your 'selfish but unhelpful' ledger system?
So, is there any way to solve the 'double spending' problem of data?
As Vana's white paper states, 'data sovereignty and collective data creation are not mutually exclusive.'
We quickly went through this white paper; a concise version could be:
The Vana protocol proposes an innovative solution by ingeniously combining privacy protection, programmable access permissions, and economic incentive mechanisms to create a brand new data economic model.
In this model, data remains encrypted at all times, and only authorized entities can access it under specific conditions; secondly, through smart contracts, data owners can precisely control who can access the data and under what conditions; more importantly, this access can be tokenized and traded, while the original data remains protected.
A more colloquial analogy could be the streaming model of the modern music industry:
Not directly selling music files (which would lead to infinite copying), but rather like Spotify's streaming service, where each use generates revenue.
Data owners do not sell data outright but retain control and can continuously earn from each use of the data. This ensures that data can be fully utilized (e.g., for AI training) while solving the 'one-time sale' issue that leads to double spending and devaluation, while the data owners always maintain complete control over their data.
Using DAO as a pool, establishing a 'data cooperative'
Specifically, what does Vana intend to do?
We can roughly divide the participants in the entire AI market into two groups—companies/AI agents that need data; individuals and organizations that (actively or passively) contribute data.
To create a higher quality AI agent, beyond public data, their demands are clear:
Access to private data, such as your health data used for healthcare AI agents
Access to paywalled data, such as paid articles and insights, used for commercial analysis AI agents
Access to closed platform data, such as more posts made by users on X, used for sentiment analysis AI agents
As the other party contributing data, intentionally or unintentionally, your demands generally include the following points:
You can access it, but the data ownership still belongs to me;
You can access it, but the data must be stored in a secure place;
You can access it, but I want to benefit from it, and pay as needed.
Traditional data usage models often place users in a passive position. For example, when AI companies need training data, they either purchase data directly from social platforms (where users cannot benefit) or need to negotiate individually with thousands of users (which is highly inefficient).
Vana's approach to solving problems is called the Data Liquidity Pool (DLP). You can understand it more down-to-earth as a 'data cooperative':
Users can centralize their data permissions in a 'pool', forming a virtual organization similar to a cooperative; this also means that the collective users have collective bargaining power while maintaining encrypted control over the original data.
Imagine a DLP composed of 100,000 Twitter users: when AI companies want to use this data, they can negotiate directly with the DLP, and the benefits will be automatically and fairly distributed among all contributors.
From the content of the white paper recently released by Vana, this data cooperative (DLP) is now being managed properly, forming four key rules:
Data norms: Membership guidelines
This is somewhat like strict membership standards, defining metadata standards, such as social media data, health data, etc.; the core is to ensure that only data that meets quality requirements is included in the pool;
Verification mechanism: Quality inspectors of the data cooperative
Assessing the quality and value of the data entering the pool to ensure that the data added is authentic, which aligns with the traditional blockchain meaning of validation nodes.
Token economics: Adjusting member behavior with rewards
Through a fair points system, incentivize quality data contributors; the more and better data contributed, the more token rewards can be obtained.
Governance rules: Cooperative charter
Regulations on how to make decisions, such as opening a new data pool, etc.; it also specifies how to handle disputes, which reflects the characteristics of DAOs that we are familiar with.
So overall, this data cooperative in the context of the crypto world is more like a DAO that makes decisions and incentives around data. The DAO manages the data pool and also determines the rules for negotiating with data users and the distribution of profits.
If you find the above statements too simplistic, then in the design of the Vana network, the aforementioned DAO model is actually operating in a serious technical manner:
Smart contract deployment. DAO creators deploy the smart contracts of the pool on the blockchain, clearly specifying how data is managed, used, and how profits are distributed.
Data preparation. Data providers prepare the data they want to contribute, which has been encrypted before being provided.
Secure storage. Data providers must first connect their wallets and prove their identity before uploading data. The uploaded encrypted data will be stored in a dedicated storage space for the contributor.
On-chain records. The system will record the access address of this encrypted data on the blockchain, ensuring that only authorized persons can access the data.
Multiple validations. Several validators will audit the data, checking its authenticity, quality, and value. These validation results will be recorded in the smart contract to ensure the credibility of the data.
Regulated usage. Verified data can be used by two types of users: machine learning researchers can pay to use data to train models; data purchasers can access data under specific conditions. All usage requires payment and strict adherence to the usage conditions specified in the smart contract.
In terms of data privacy protection, due to space and technical knowledge limitations, I will not elaborate further here.
If you are concerned about whether this data might leak, just grasp this main line: all personal data in Vana always remains encrypted, just like being placed in a safe where the user holds the key. Even if it needs to process this data, it can only be done in a special secure environment (TEE), similar to a bank's special clearing room, where all operations are strictly monitored and recorded.
It is particularly noteworthy that the system achieves flexible but secure access control through the combination of smart contracts and cryptographic mechanisms. It can control who can access what data at what time, and all access records will be properly preserved for auditing.
Using DAO as a data pool, a data cooperative model can protect personal data sovereignty and benefits while also allowing AI agents that need more personalized data to make full use of it.
A flourishing variety of data DAOs, each specializing in their own areas
Currently, the data liquidity pools on Vana are not just at the planning stage; they have indeed formed various data DAOs. Data in each DAO is aimed at a specific vertical scenario for different AI needs.
Taking the Volara DAO, which focuses on X (Twitter), as an example, you can connect your Twitter to this platform, and then upload all your tweets and related social data. Volara DAO will reward you with tokens from this DAO based on your contributions.
Note that the direct rewards are not Vana, but the DAO's own tokens, such as $VOL.
This is very similar to the current popular Virtuals, where a base currency has corresponding tokens created by different projects. Holding VOL qualifies you for $VANA airdrops, and the nested asset model creates space for more gameplay.
We have organized 16 currently popular data DAOs in Vana and made a detailed classification of them.
For ordinary players, this feels more like a concept of 'data mining'—if you have faith in a certain DAO, you can contribute data according to its rules, and in return, you will receive corresponding rewards and airdrops.
However, you don't necessarily have all the data, so you also need to look at what data you can contribute according to the classifications below, and find the best way to maximize your benefits:
Platform-based data DAOs
Device and data generation DAOs
Human insights and financial DAOs
Health DAOs
Overall, since the developer testnet launched in June 2024, the Vana network has attracted 1.3 million users, over 300 data DAOs, and 1.7 million daily trading volume.
With the launch of the mainnet and the issuance of tokens, perhaps we will see more data DAOs emerge under the impetus of economic incentives.
Dual-layer tokens are more in line with the gameplay of the version.
You may have already noticed that the above DAOs all have their own sub-tokens and corresponding links with the base currency VANA (such as airdrops).
This involves a carefully designed dual-layer token economic model.
Imagine the traditional data market: healthcare data, financial data, social data—their value standards and usage scenarios vary greatly. Measuring and incentivizing such diverse data contributions with a single token is like using a single ruler to measure everything—from planets to atoms. This is clearly not precise enough or flexible enough.
VANA adopts a more elegant solution: setting a unified base token (VANA) at the protocol level while allowing each data DAO to issue its own exclusive tokens.
Both the parent and child tokens have different roles and functions:
VANA:
Supply of 120 million tokens. First, it ensures the corresponding network security by requiring validators to stake VANA;
Secondly, it serves as the basic payment currency for all transactions. For example, if AI companies want data from this DAO, they need to pay with VANA;
Most importantly, it requires each data DAO to stake at least 10,000 VANA to operate, serving as a 'good faith deposit' to ensure the long-term commitment of DAO operators to the ecosystem.
Tokens of data DAOs:
Each data DAO can design a token economic model that suits its field characteristics. For instance, a healthcare data DAO might emphasize the integrity and accuracy of data, thus designing special reward mechanisms to encourage high-quality patient data contributions; while a social data DAO might focus more on user interaction activity and influence.
These exclusive tokens are not just simple points but build a complete value capture system: when data is used, both VANA and the corresponding DAO tokens must be paid. It's like paying a 'venue fee' (VANA) and a 'special service fee' (DAO token) when using data.
Does this gameplay remind you of Virtuals?
Similarly, the brilliance of the dual-layer token system lies in creating a self-sustaining economic cycle: using data requires consuming tokens, some of which will be burned, causing deflationary pressure; at the same time, high-quality data contributions will earn new token rewards, providing a moderate inflationary incentive. This balance ensures the stability of token value and encourages continuous data contributions.
Vana serves as the base currency, with gas and staking functions. Each sub-DAO issues its own token, paired with the base currency VANA, allowing the base currency to capture the benefits of ecological prosperity.
From the perspective of creating assets and increasing asset efficiency, the VANA model is clearly aligned with the current AI agent craze.
For individuals, this system turns data into a truly sustainable asset. Data providers no longer sell data outright but continuously share the benefits brought by data usage through holding tokens. This is akin to moving from a 'buyout system' to a 'royalty sharing system', greatly improving the interests of data creators.
At the same time, as the Vana mainnet will soon launch (the token economics have been announced, and the mainnet is in the pre-launch phase), after understanding this dual-token gameplay, there are at least two things you can participate in:
First, as mentioned above, contribute data to different data DAOs in hopes of obtaining sub-DAO tokens and corresponding VANA airdrops; the summary link is here.
Second, with the launch of the mainnet, we also noticed changes on Vana's official website. Currently, a new datahub page has been added to manage the different data DAOs you participate in and their corresponding tokens.
Currently, there is a pre-registration activity on this page to prepare to associate your identity and earn rewards; interested players are advised to lay out their plans in advance.
After completing this registration, you will be prompted to become an 'Early Explorer'.
Summary
In the current AI Agent hotspot, the influence of AI Agents is growing, until it fills your information stream and investment list.
However, the narrative of Vana actually states that your own influence is greater than you might think.
By contributing various types of data, you become part of the AI craze; and through the tokenization of data assets, you also gain another way to create around assets.
It cannot be denied that in the crypto world, creating assets is a clear path. Those closer to assets can gain more narrative space and benefits.
In fact, when your data can be tokenized, I believe this is a hidden line that fits the clear line, and it is also a key puzzle piece for individuals to embrace, utilize, and participate in the trend of AI intelligence.
The narrative at the data layer has yet to be fully developed; whether Vana will be discovered for its value remains to be seen.
