The Fusion of AI and Web3: Who Will Lead the New Era of Chain-Level Infrastructure Revolution?

When a technological paradigm shift truly occurs, we often see the hype first, rather than the system. The AI wave we are experiencing is no different.

As a primary investor, I always believe that betting on the deep transformative forces of the industry is far more valuable than chasing superficial narratives.

In the past year, I have been exposed to a large number of projects related to RWA, consumer applications, and information finance. They are undoubtedly exploring the intersection between the real world and on-chain systems. However, an increasingly obvious trend is that no matter which route the project takes, it ultimately needs to enter the collaborative logic of AI, utilizing AI to enhance competitiveness and efficiency.

For example, RWA projects need to consider how to use AI for risk control optimization, off-chain data verification, and dynamic pricing; consumer applications and DeFi projects require AI for user behavior prediction, strategy generation, and incentive distribution. Similar needs exist in other fields.

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 have the capacity for AI integration and support, it will not be able to sustain the complex collaboration of the next generation of applications.

In my opinion, the future of AI is not just about "becoming stronger" and "being used more widely." 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 for the first time to participate in content creation and turn ideas into products, thereby giving rise to an entire open ecosystem.

AI is also following this path: intelligent agents will become smart co-creation entities for everyone, helping you transform expertise, creativity, and tasks into automated productivity tools, even enabling monetization. This is a question that the current Web2 world struggles to answer, and it reflects some underlying logic I focus on 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 that is currently trying to systematically build the underlying operation of AI from a chain-level structure: Sahara.

The essence of investment is a worldview that recognizes the value system of choices.

My investment logic is not simply about combining public chain narratives with AI and then betting on teams that look good on paper.

Investing is essentially a choice of worldview. I have always been asking a core question: Can the future of AI be collectively owned by more people? Can it leverage blockchain to reconstruct the logic of value attribution and distribution in AI, allowing ordinary users, developers, and other different roles to have opportunities to participate, contribute, and continuously benefit? Only with the emergence of this logic do I believe that such projects have the potential to be disruptors, rather than "just another useless public chain."

In order to find the answer, I researched almost all the AI projects I could access until I encountered Sahara. The co-founder of Sahara told me: to build an open, participatory ecosystem that everyone can own and benefit from.

This sentence is simple, but it precisely hits the soft spot of traditional public chains: they often serve developers in a one-sided manner, and the design of token economics is mostly limited to Gas fees or governance, rarely able to truly support a positive cycle of the ecosystem, making it even harder to sustain the development of an emerging track.

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.

The true paradigm shift lies not in creating a single product, but in building a supportive system. And Sahara is one of the cases I anticipated the most at that time.

From Investment to 8x Valuation Follow-up Heavy Investment

If I initially invested in Sahara because it is realizing the true mission of AI in my heart - building an AI economy and infrastructure system. Then what made me rush to invest again at an 8x pre-round valuation in just half a year is the rare strength I felt in this team.

Two co-founders, one of whom is the youngest tenured professor at the University of Southern California, specializing in AI. The value of a tenured professor in American universities born in the 90s is not only reflected in the academic field but also in the fact that this age group still has dreams, energy, and the determination to achieve those dreams. Having known Professor Ren for over a year, I have witnessed what it means to work more than ten hours a day, with stable emotions and humility.

Another co-founder, Tyler, who previously served as the investment director at a well-known blockchain lab, was responsible for North American investments and incubators, so his understanding of Web3 goes without saying. His level of self-discipline is astonishing: he only sleeps in multiples of 1.5 hours, insists on working out regardless of how busy he is to maintain his condition, and completely avoids sugar to keep his mind clear, working over 13 hours a day. I once joked that he was like a robot, to which he simply replied: "I'm lucky to have this level of busyness today." He derives satisfaction from advancing project progress; dream-making is his passion, and he requires no other motivation.

I am very glad to have met them, and it has also changed me. I have started to regulate my sleep as much as possible, my emotions have gradually stabilized, and I have started working out...

Whenever someone says that Sahara gained capital favor due to luck, I always candidly add, "The pursuit of capital is an inevitable result." I vividly remember the scene where, despite the difficulties in primary market financing during this round, Sahara was chased by investors for investment.

What everyone remembers is that some well-known investment institutions have invested in Sahara. Sahara has opened the investment era for Samsung's entry into the Web3 AI field, and its winning of the Samsung AI Award is an important reason for Samsung's investment. In addition, some funds that focus on AI and national banks have also shown favor towards Sahara. You can see a group of institutions that are more focused on traditional technology and industrial resources starting to quietly bet on AI × Web3 because of Sahara.

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

This is also 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 (millions in the queue). Their clients 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 "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 combining 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 labeled data 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 labeling, which not only increases costs but also diminishes the value-driven nature 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 also invests in the future with data.

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

A typical example is a voice synthesis project on a certain blockchain, which quickly built a high-quality dataset covering multiple languages and accents by leveraging Sahara's decentralized data collection and human-machine collaborative annotation, significantly 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 model 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 protection of all participants' rights 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 considered a scenario with long-term value support?

Imagine if you want to build an AI application, you would 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 with 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 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 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 phases. Sahara is building 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, and computing power contribution nodes can continuously benefit at different stages, rather than relying solely 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 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 data plunder and model buyouts; it cannot just benefit a few individuals. It must be open, collaborative, and mutually beneficial—where everyone can participate, every valuable contribution can be recorded, and continuous returns can be obtained in the future.

But the closer to the real structure, the more challenges there are.

Although I am optimistic about Sahara, I will not hide 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 has been designed from the beginning to be open, full-chain, and standardized: it supports deployment on any EVM-compatible chain, and also provides a standard API interface, allowing Web2 systems—whether it is an e-commerce backend, enterprise SaaS, or mobile app—to directly call Sahara's model services and complete on-chain settlement.

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

To become a trusted, adopted, and integrated AI protocol layer, Sahara's key lies in how ecological 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: serving several leading technology companies and research institutions by providing them with relevant data services and addressing some of the industry's most challenging data demand issues, thus validating this system.