The Integration of AI and Web3: Who Will Lead the Infrastructure Development of the New Era?

The true transformation of technological paradigms often manifests first as a craze, rather than a complete system. The current wave of AI is a prime example of this.

As a primary investor, I always believe that the deepest transformative power in the investment industry is far more valuable than chasing superficial narratives.

In the past year, I have come across a large number of projects in RWA, consumer, and information finance, all of which are exploring the intersection of the real world and blockchain systems. However, an increasingly obvious trend is that regardless of the project direction, it ultimately needs to incorporate AI collaborative logic to enhance competitiveness and efficiency.

For example, RWA projects need to consider how to utilize AI for risk control optimization, off-chain data verification, and dynamic pricing; consumer or DeFi projects require AI to achieve functions such as user behavior prediction, strategy generation, and incentive distribution.

Therefore, whether it is asset digitization or user experience optimization, these seemingly independent narratives will ultimately converge on the same technological logic: if the infrastructure does not have the capability to integrate and support AI, it cannot sustain the complex collaboration of the next generation of applications.

In my opinion, the future of AI is not just about becoming "stronger and stronger" and "more widely used". The real paradigm shift lies in the reconstruction of collaborative logic. Just like the early transformations of the internet, it was not because we invented DNS or browsers, but because it allowed everyone to participate in content creation for the first time, turning ideas into products, thereby giving rise to an entire open ecosystem.

AI is also on this path: intelligent agents will become co-creation partners for everyone, helping to transform expertise, creativity, and tasks into automated productivity tools, and even realize value monetization. This is a question that the current Web2 world struggles to answer, and it is also the underlying logic of my interest in the AI + Web3 track: making AI collaborative, transferable, and profit-sharing is the system that is truly worth building.

Today I want to discuss the only project currently attempting to systematically build the underlying operations of AI from a blockchain-level structure: Sahara.

The essence of investment is a choice of worldview

My investment logic is not simply to combine the public chain narrative with AI, and then choose teams with better backgrounds to bet on. Investment is essentially a choice of a certain worldview, and I have been exploring a core question: Can the future of AI be owned by more people collectively?

Can it use blockchain technology to reconstruct the value attribution and distribution logic of AI, allowing ordinary users, developers, and other roles to have the opportunity to participate, contribute, and continuously benefit? Only when this logic appears do I believe that relevant projects have the potential to become disruptors, rather than "just another useless public chain."

In order to find answers, I almost examined all the AI projects I could access until I encountered Sahara. The answer given by Sahara's co-founder Tyler is: to build an open, participatory ecosystem that everyone can own and benefit from.

This simple answer precisely hits the Achilles' heel of traditional public chains: they often serve developers in a one-dimensional way, and the token economic design is mostly limited to Gas fees or governance, rarely able to truly support a positive cycle of the ecosystem, and even more difficult to sustain the development of emerging tracks.

I fully understand that this path is filled with challenges, but precisely because of that, it is a revolution that cannot be refused—this is also the reason for my firm investment.

As I emphasized when discussing the "evolution from Web2 to Web3": the real paradigm shift is not about building a single product, but about creating a supportive system. And Sahara is one of the most anticipated cases in my earlier predictions.

From investment to 8x valuation follow-up investment

If I initially invested in Sahara because it is fulfilling my vision of a true leader in AI – building an AI economy and infrastructure system, then what drove me to eagerly invest at an 8-fold valuation within just six months is the extremely rare strength I felt in this team.

The two co-founders have demonstrated extraordinary talent and perseverance. One of them is the youngest tenured professor at the University of Southern California, specializing in the field of AI. The other, Tyler, previously served as the Director of Investments at a well-known trading platform's lab, responsible for North American investments and incubators, with a deep understanding of Web3. Their discipline and level of commitment are astounding, working over 13 hours a day and dedicating all their energy to the project.

It is precisely this kind of team that has garnered Sahara the favor of the capital market. Not only are well-known investment institutions involved, but it has also attracted some organizations that are more focused on traditional technology and industrial resources to start paying attention to the integration of AI and Web3.

These all reflect the positive recognition from the market regarding the depth of Sahara technology, the team's background, and the system design and execution capabilities. The project has also demonstrated some substantial progress: the testnet has activated over 3.2 million accounts, there are over 200,000 data platform annotators, the clients served include several tech giants, and it has achieved revenue at the level of tens of millions of dollars.

On this infrastructure chain, at least from "who will do it" to "can it be done", Sahara has gone deeper and more steadily than 99% of "AI concept projects".

The Ultimate Challenge of Public Blockchains: Ensuring Continuous Benefits for All Contributors and Driving Positive Economic Cycles

Returning to the initial judgment logic: In a system where AI and blockchain are combined, is there really a mechanism that allows every contributor to be seen, recorded, and continuously rewarded?

Model training and data optimization rely heavily on a large amount of labeling and interaction support; conversely, if user contributions are lacking, the project will have to invest more funds to purchase data and outsource labeling, which not only increases costs but also weakens the value-driven aspect of community co-construction.

Sahara is one of the few Web3 AI projects that allows ordinary users to "participate in data construction from day one." Its data annotation task system operates daily, with a large number of community users actively participating in annotations and prompt creation. This not only helps to improve the system but also invests in the future with data.

Through the mechanism of Sahara, not only has the model quality been improved, but it has also allowed more people to understand and participate in this decentralized AI ecosystem, linking data contributions with rewards to form a true positive feedback loop.

For example, a certain project quickly built a high-quality dataset covering multiple languages and accents through Sahara's decentralized data collection and human-machine collaborative annotation, significantly improving the training efficiency of its speech-related models. This has garnered a lot of attention and usage for its open-source project. At the same time, users participating in data annotation also received token rewards, creating a two-way incentive loop between developers and data contributors.

Sahara's "permissionless copyright" mechanism ensures the protection of the rights of all participants while guaranteeing the open circulation and reuse of AI assets—this is the underlying logic driving the explosive growth of the entire ecosystem.

Why is this a scenario with long-term value support?

Imagine if you were to build an AI application, you would naturally want your model to be more accurate and closer to real users than others.

The key advantage of Sahara is that it connects you to a large and active data network—hundreds of thousands, potentially millions of annotators in the future. They can continuously provide you with customized, high-quality data services, allowing your models to iterate faster.

More importantly, this is by no means a one-time transaction. Through Sahara, you are connecting to a potential community of early users; and these contributors are likely to become real users of your product in the future.

This connection is not a one-time buyout; through Sahara's smart contract system and rights confirmation mechanism, it has achieved a long-term, traceable, and sustainable incentive system. Regardless of how many times the data is accessed, contributors will receive continuous profit sharing, with earnings dynamically linked to usage behavior.

But this is not just a revenue model for data labeling and model training stages. Sahara builds an economic system that covers the entire lifecycle of AI models, with a built-in profit-sharing mechanism at every stage after the model goes live, including calls, combinations, and cross-chain reuse, allowing value to be captured over a longer period.

Model developers, optimizers, validators, computing power contribution nodes, etc. can now continue to benefit at different stages, rather than just relying on one-time transactions or buyouts.

This system brings a compound effect for model combination calls and cross-chain reuse. A trained model, like building blocks, can be repeatedly called and combined by different applications, with each call generating new revenue for the original contributor.

For this reason, I agree with Sahara's underlying belief: a truly healthy AI economic system cannot be just about data plunder and model buyouts, nor can it be about a few people reaping all the benefits. It must be open, collaborative, and mutually beneficial—where everyone can participate, every valuable contribution can be recorded, and continuous rewards can be obtained in the future.

The Challenge of Approaching Real Structures

Although I am optimistic about Sahara, I will not ignore the challenges the project will face because of my investment position.

One of the major advantages of the Sahara architecture is that it is not limited to any particular chain or single ecosystem. Its system was designed from the very beginning to be open, cross-chain, and standardized: it supports deployment on any EVM-compatible chain, while also providing standard API interfaces that allow Web2 systems—whether it's e-commerce backends, enterprise SaaS, or mobile apps—to directly call Sahara's model services and complete on-chain settlements.

However, despite the extreme scarcity of this architectural design, it also carries a core risk: the value of the infrastructure lies not in "what it can do", but in "who is willing to do what based on it".

To become a trusted, adopted, and integrated AI protocol layer, the key for Sahara lies in how ecosystem participants assess its technological maturity, stability, and future predictability. Although the system itself has been built, whether it can truly attract a large number of projects to implement based on its standards remains uncertain.

It is undeniable that Sahara has achieved key validation: providing relevant data services to several leading technology companies and addressing some of the industry's most challenging data demand issues, becoming an early signal of the feasibility of this system.

However, it needs to be noted that these collaborations mainly come from the Web2 world. What truly determines the long-term development of Sahara is still the maturity and penetration of the entire Web3 AI sector. Sahara benefits from the major trend of Web3 AI, but in order to truly unlock the value of its infrastructure, it still requires more Web3 native AI products and technical solutions to be implemented and perfected.

But don't forget, Sahara is currently "one of a kind".

In the native chain-level infrastructure track designed for AI, although there are many imitators proposing conceptual frameworks, only Sahara has managed to fully implement the technical closed loop and real revenue from on-chain rights confirmation, off-chain execution, and cross-chain invocation, and has received validation from actual customers.

This not only brings Sahara a "monopoly advantage", but also introduces structural risks: once successful, it will define the industry benchmark for the entire Web3 × AI infrastructure; but if it fails, AI Layer1 may be seen as a premature layout.

Since it is now the only option in this field, the market's judgment on it is naturally going to be more stringent and rational - it must withstand the tests of time and ecology.

For All Builders and Observers

For me, the core of every primary investment decision revolves around three things: the depth of understanding of the world, the dimensions of trend judgment, and the willpower of the team to navigate through cycles. Products and functions are certainly important, but they often only reflect these underlying cognitions.

Web3 is not short of ideas, nor stories; what it lacks is the hand to turn logic into order, and it lacks people who truly know what to hold on to and what to give up.

I cannot guarantee that Sahara can become the next paradigm-level chain. But it is indeed the only attempt worth taking seriously, observing carefully, and betting on at the moment.