The Integration of AI and Web3: Who Will Lead the Next Generation of Infrastructure Development?

The true shift in technological paradigms is often accompanied by a frenzy rather than a complete system. The current AI wave is no exception. As a first-level investor, I have always believed that focusing on deep industry transformations is more valuable than chasing superficial narratives.

In the past year, I have examined numerous projects such as RWA, Consumer, InfoFi, etc., all of which are exploring the intersection of the real world and on-chain systems. However, one obvious trend is that regardless of the project's direction, they ultimately need 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 user behavior prediction, strategy generation, and incentive distribution. This indicates that, whether it is asset digitization or experience optimization, these seemingly independent narratives will ultimately converge into the same technological logic: the infrastructure must have the integration and carrying capacity of AI to support the complex collaboration of the next generation of applications.

The future of AI is not just about becoming more powerful or being used more widely; the real paradigm shift lies in the reconstruction of collaborative logic. Just like the early transformations of the internet were not merely about inventing DNS or browsers, but rather about allowing everyone to participate in content creation and turning ideas into products, thus giving rise to an entire open ecosystem.

AI is heading down a similar path: Agents will become intelligent co-creators for everyone, assisting in transforming expertise, creativity, and tasks into automated productivity tools, even enabling value monetization. This is a problem that the current Web2 world struggles to solve, and it is also the underlying logic of my focus on the AI+Web3 track: making AI collaborative, transferable, and revenue-sharing is the system that is truly worth building.

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

The Essence of Investment: The Choice of Worldview and Value System

My investment logic is not simply combining the public chain narrative with AI and then choosing a well-established team. Investment is essentially a choice of worldview, and I have been pondering 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 different roles such as ordinary users and developers to participate, contribute, and continuously benefit? I only believe that related projects have the potential to be disruptors when this logic emerges, rather than just "another public chain".

In the process of seeking answers, I came across numerous AI projects until I encountered Sahara. The vision presented by Sahara's co-founder Tyler is to build an open, participatory ecosystem that everyone can own and benefit from. This simple answer precisely points out the shortcomings of traditional public chains: they often serve developers in a one-way manner, and the design of token economies is mostly limited to Gas Fees or governance, making it difficult to truly support the positive cycle of the ecosystem and even harder to sustain the development of emerging tracks.

I am well aware that this path is full of challenges, but precisely because of this, it is a revolution that cannot be refused, and it is also the reason for my firm investment. A true paradigm shift is not about creating a single product, but about building a supportive system. Sahara is one of the cases I look forward to the most.

From investment to eightfold valuation follow-up heavy investment

I initially invested in Sahara because it meets my expectations for a leading AI project - to build an AI economy and infrastructure system. What prompted me to increase my investment by 8 times the previous round valuation in just half a year was the rare strength I felt in this team.

The two co-founders each have their own unique qualities: one is the youngest tenured professor at the University of Southern California, specializing in the field of AI. His academic achievements are reflected not only in his expertise but also in his ability to maintain the energy and determination to realize his dreams at such a young age. Having known Professor Ren for over a year, I have witnessed his genius demeanor, working over ten hours a day, maintaining emotional stability, and remaining humble.

Another Tyler, who served as the investment director of a well-known exchange's lab, is responsible for North American investments and incubators, and has a profound understanding of Web3. His self-discipline is astonishing: sleep time is precisely controlled, he insists on working out regardless of how busy he is, and he avoids sugar to keep his mind clear, working over 13 hours a day. I once jokingly called him a robot, and he replied calmly, "I am lucky to have the busy life I have today." He derives satisfaction from advancing project progress; dreaming is his passion, requiring no other motivation.

Meeting them has changed my lifestyle; I have also started to maintain a regular schedule, my emotions are more stable, and I have begun to work out...

Therefore, when someone says that Sahara gained capital favor due to luck, I will unabashedly point out that "the capital's pursuit is a necessary result." In this round of difficulty in primary market financing, Sahara is being chased by investments in the primary market, which is an impressive scene.

In addition to the participation of well-known investment institutions, Sahara has also ushered in an investment era for a major technology company entering the Web3 AI field, and its receipt of the company's AI award is an important factor in facilitating this investment. Moreover, some AI-focused funds and national-level banks have also become supporters of Sahara. It can be seen that a group of institutions that are more inclined towards traditional technology and industrial resources are quietly laying out in the AI × Web3 field because of Sahara.

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

This also explains why Sahara is able to exhibit some substantial structural indicators:

More than 3.2 million accounts have been activated on the test network, with over 200,000 data platform annotators (millions are in line). Their clients include several leading technology companies, and they have achieved revenue in the tens of millions of dollars.

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

The Ultimate Challenge of Public Chains: Achieving Continuous Benefits for All Contributors and Driving a Positive Economic Cycle

Returning to our initial judgment logic: In the combined system of 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 heavily on a large amount of annotation and interaction support; conversely, if there is a lack of user contributions, the project will have to invest more funds to procure data and outsource annotation, which not only increases costs but also undermines 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 labeling task system operates daily, with a large number of community users actively participating in labeling and prompt creation. This not only helps improve the system but is also an investment in the future through data.

Through the mechanism of Sahara, not only has the quality of the model been improved, but it also allows more people to understand and participate in this decentralized AI ecosystem, linking data contributions with rewards, thus forming a true virtuous cycle.

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

At the same time, users participating in data annotation also receive token rewards distributed 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 safeguarding the rights and interests of all participants—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 vast 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 model iteration speed to be 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 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 right confirmation mechanism, it can achieve a long-term, traceable, and sustainable incentive system.

Regardless of how many times the data is called, 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 constructs an economic system that covers the entire lifecycle of AI models, with built-in profit-sharing mechanisms at every stage, including model deployment, invocation, combination, and cross-chain reuse, allowing value to be captured over a longer period.

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

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 creating new benefits for the original contributor.

For this reason, I agree with Sahara's underlying belief: a truly healthy AI economic system cannot be just about the plunder of data and the buyout of models; it cannot simply allow a few people 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 rewards can be gained in the future.

The closer we get to the real structure, the more challenges there are.

Although I am optimistic about Sahara, I will not overlook the challenges that the project will face due to my investment position.

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

Its system was designed from the beginning to be open, full-chain, and standardized: supporting 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 presents 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 ecological participants assess its technology maturity, stability, and future predictability. Although the system itself has been built, whether it can truly attract a large number of projects to land based on its standards remains uncertain.

Undoubtedly, Sahara has achieved key validation: serving several leading technology companies by providing them with relevant data services and addressing some of the industry's most challenging data demand issues, thus proving the feasibility of this system.