AI × Web3: Who will build the foundational chain for this era?

The true transformation of technological paradigms often manifests first as a surge, rather than a complete system. The AI wave we are currently experiencing is no exception.

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

In the past year, I have come across numerous projects exploring the integration of real-world systems with on-chain systems, such as RWA, Consumer, and infoFi. An increasingly evident trend is that, regardless of the project's direction, there is a need to incorporate AI collaboration logic to enhance competitiveness and efficiency.

For example, RWA will need to consider how to utilize AI for risk control optimization, off-chain data verification, and dynamic pricing in the future. Consumer or DeFi projects also require AI for user behavior prediction, strategy generation, and incentive distribution.

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

I believe that the future of AI is not just about becoming more powerful or being used more widely, but rather a true paradigm shift lies in the reconstruction of collaborative logic. Just like the early transformation of the internet, it wasn't just because of the invention of DNS or browsers, but because it allowed everyone to participate in content creation for the first time, turning ideas into products, thus giving rise to an entire open ecosystem.

AI is following a similar path: Agents will become intelligent co-creation entities for everyone, helping to transform expertise, creativity, and tasks into automated productivity tools, even enabling monetization. This is a problem that the current Web2 world struggles to address, and it is part of the fundamental logic I focus on in the AI+Web3 space: making AI collaborative, transferable, and profit-sharing is the system that is truly worth building.

Today, I want to discuss the only project so far that attempts to systematically build the underlying framework for AI operations starting from the chain-level structure: Sahara.

The essence of investment is the worldview, affirming the value system of choices.

My investment logic is not simply to combine the narrative of public chains with AI, and then choose teams with a good background to bet on.

Investment is essentially a choice of worldview. I have been thinking about a core question: Can the future of AI be jointly owned by more people? Can it leverage blockchain to reconstruct the value attribution and distribution logic of AI, allowing ordinary users, developers, and other roles to have opportunities 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 abandoned public chain."

In search of answers, I have almost studied all the AI projects I could access, until I came across 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-sided manner, and the token economic design is mostly limited to Gas Fees or governance, making it difficult to truly support a positive cycle for the ecosystem, and even harder to sustain the development of emerging sectors.

I am well aware that this path is full of challenges, but precisely because of that, it is a revolution that cannot be refused — and that is my reason for investing steadfastly.

As I emphasized earlier in my discussion on the "evolution from Web2 to Web3": the real paradigm shift lies not in creating a single product, but in building a supportive system. Sahara is one of the cases I anticipated the most at that time.

From Investment to 8x Valuation Follow-on Investment

I initially invested in Sahara because it is pursuing what I believe to be the true leading mission of AI — building the AI economy and infrastructure systems. What made me rush to invest at 8 times the previous round valuation in just half a year is the rare strength I felt in this team.

Among the two co-founders, one is the youngest tenured professor at the University of Southern California, specializing in AI. The value of a tenured professor in their 90s in American universities is not only reflected in the academic field, but more importantly, this age still has dreams, energy, and the determination to achieve those dreams. Having known Professor Ren for more than a year has shown me what it means to work more than ten hours a day, with stable emotions and humility.

Tyler is the former North American Investment Director of a well-known investment institution, responsible for North American investments and incubators, and his understanding of Web3 goes without saying. His self-discipline is astonishing: he only sleeps in multiples of 1.5 hours, insists on working out to maintain his condition no matter how busy he is, and completely avoids sugar for a clear mind, working over 13 hours a day. I once jokingly called him a robot, to which he simply responded, "I am lucky to have this busy life today." His source of dopamine comes from making progress on projects every day; dreaming is his passion, and he doesn't need any other fuel.

Meeting them changed me. I also started to maintain a regular schedule as much as possible, my emotions gradually stabilized, and I began working out...

So when someone says that Sahara gained the favor of capital due to luck, I will candidly add, "The pursuit of capital is an inevitable result." I vividly remember the difficulties of primary financing in this market round, but Sahara was being chased for investment in the primary market.

What everyone remembers is that certain well-known investment institutions have invested in Sahara. Sahara has ushered in an investment era for a large technology company entering the Web3 AI field, and its receipt of the company's AI award is a significant reason for the investment. In addition, some AI-heavy funds, national banks, and others are also prominent guests of Sahara. You can see a group of institutions that are more focused on traditional technology and industrial resources beginning to quietly bet on AI × Web3 because of Sahara.

Capital will only pay for a clear direction and execution capability - this is a positive feedback on the depth of Sahara technology, team background, system design, and execution capabilities.

This also explains why it can produce some real and solid structural indicators:

Over 3.2 million accounts have been activated on the test network, with more than 200,000 data platform annotators (with millions more waiting in line). The clients they serve include several leading technology companies, and they have already achieved revenue in the 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 the "AI Narrative projects".

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

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

Model training and data optimization rely on a large amount of annotated and interactive support; conversely, if there is a lack of user contributions, the project itself has to invest more funds to procure data and outsource annotation, which not only increases costs but also weakens the value-driven 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 involved in annotation and prompt creation. This not only helps improve the system but also invests in the future with data.

Through the mechanism of Sahara, not only is the model quality improved, but more people are also made aware of and involved in this decentralized AI ecosystem, linking data contributions to benefits and forming a true positive cycle.

A typical example is a project on a public chain that quickly built a high-quality dataset covering multiple languages and accents by leveraging Sahara's decentralized data collection and human-machine collaborative labeling, greatly improving the training efficiency of its TTS and voice cloning models. This also propelled its open-source project to receive thousands of GitHub stars and over 2 million downloads.

At the same time, users participating in data annotation also received token rewards issued by the project, forming a two-way incentive loop between developers and data contributors.

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

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

Imagine if you want to build an AI application, you naturally hope that your model is more accurate and closer to real users than others.

The key advantage of Sahara is that it connects you to a vast and active data network—hundreds of thousands, and in the future, millions of annotators. 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 early user community; and these contributors are very likely to become the actual 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 enables 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 of model deployment, invocation, combination, and cross-chain reuse, allowing value to be captured over a longer period.

Model developers, optimizers, validators, and computing power contributing nodes can now continuously benefit at different stages, rather than just relying on a one-off transaction or buyout.

Such a 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 merely about the plunder of data and the buyout of models, nor can it be about allowing a select few to reap all the benefits. It must be open, collaborative, and mutually beneficial—where everyone can participate, every valuable contribution can be recorded, and continuous returns can be gained in the future.

But the closer we get to the true structure, the more challenges there are.

While I am optimistic about Sahara, I will not cover up the challenges that the project will face because of my investment stance.

One of the major advantages of the Sahara architecture is that it is not limited to any specific chain or single ecosystem.

Its system was designed from the beginning to be open, full-chain, and standardized: it supports deployment on any EVM-compatible chain, while also providing standard API interfaces that allow Web2 systems—whether 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 does not lie 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 ecological participants assess its technological maturity, stability, and future predictability. Although the system has been built, it remains uncertain whether it can truly attract a large number of projects to be implemented based on its standards.

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

However, it should be noted that these collaborations mainly come from the Web2 world. The long-term development of Sahara is still determined by the maturity and penetration of the entire Web3 AI sector. Sahara benefits from the overarching trend of Web3 AI, but to truly unlock the value of its infrastructure, it still requires the implementation and improvement of more Web3 native AI products and technical solutions.

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