AI × Web3: The Infrastructure Revolution of a New Era

The true transformation of technological paradigms often precedes the emergence of systems. The wave of AI that we are currently experiencing is no exception.

As a primary investor, I always believe that the transformative power deep within the industry is more valuable than chasing surface narratives. Over the past year, I have examined numerous projects in RWA, Consumer, infoFi, etc., all of which are exploring the intersection of the real world and on-chain systems.

But the trend is becoming increasingly clear: regardless of the project's direction, it ultimately needs to integrate AI collaborative logic to enhance competitiveness and efficiency. For example, RWA requires AI for risk control optimization, off-chain data verification, and dynamic pricing; Consumer and DeFi need AI to perform 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: infrastructure that does not possess AI integration and carrying capabilities cannot support the complex collaboration of the next generation of applications.

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

AI is also heading down this path: Agents will become intelligent co-creators for everyone, helping to 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 is also the underlying logic of my focus on 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 chain-level structure: Sahara.

The essence of investment is the worldview, recognizing the value system of choice.

My investment logic is not simply to combine public chain narratives with AI and then choose to bet on teams that seem to have a good background.

Investment is essentially a choice of worldview, and I have always been questioning a core issue: Can the future of AI be collectively 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 the opportunity to participate, contribute, and continuously benefit? Only when this logic emerges do I believe such projects have the potential to be disruptors, rather than just "abandoned public chains +1".

In search of answers, I examined almost all accessible AI projects 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 sentence is concise yet strikes at the soft spot of traditional public chains: they often serve developers in a one-dimensional way, and the design of their token economies is mostly limited to Gas Fees or governance, making it difficult to truly support a positive cycle in the ecosystem, and even harder to sustain the development of emerging tracks.

I fully understand that this road is full of challenges, but precisely because of this, it is an irresistible revolution - and that is the reason for my firm investment.

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

From Investment to 8x Valuation Follow-on Investment

If I initially invested in Sahara because it was doing what I believe a true AI leader should be doing—building an AI economy and infrastructure system, then what made me rush to invest again at eight times the pre-round valuation in just half a year is the extremely rare strength I felt in this team.

Among the two co-founders, one is the youngest tenured professor at USC, 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 group still possesses dreams, energy, and the determination to realize those dreams. Having known Professor Ren for more than a year, I have witnessed what it means to work for over ten hours a day, while being emotionally stable and humble.

Another Tyler is the former North America Investment Director of a well-known investment institution, responsible for investments and incubators, and 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 exercising no matter how busy he is to maintain his condition, and refrains from alcohol to keep his mind clear, working over 13 hours a day. I once joked that he is like a robot, and he simply responded calmly: "I am lucky to have this busy life today." His source of dopamine comes from advancing project progress 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 routine as much as possible, my emotions gradually stabilized, and I began to work out...

So when someone says that Sahara gained capital favor due to luck, I always unreservedly add, "The pursuit of capital is an inevitable result." I vividly remember the difficulty of primary financing in this market cycle, yet Sahara was being chased by investors in the primary market.

What everyone remembers is that some well-known investment institutions have invested in Sahara. Sahara has opened the investment era for a certain large technology company to enter the Web3 AI field, and its winning of the company's AI award is an important 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 certain direction and execution capability—this is a 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:

More than 3.2 million accounts have been activated on the test network, with over 200,000 data platform annotators. Millions are queued (, and their clients include several leading technology companies, having 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".

![AI × Web3: Who will build the chain for this era?])https://img-cdn.gateio.im/webp-social/moments-cf477a069888f7e6e1a56aa0d7248930.webp(

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 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 on substantial support from annotations and interactions; conversely, if there is a lack of user contributions, the project itself has to invest more funds to procure data and outsource annotations, which not only increases costs but also diminishes 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 participating 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 has the quality of the model been improved, but it has also allowed more people to understand and participate in this decentralized AI ecosystem, linking data contributions with rewards, thus forming a true virtuous cycle.

A voice project on a certain blockchain is a typical example. 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 led to its open-source project receiving 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 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 it said that this is a scenario with long-term value support?

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

The key advantage of Sahara is that it connects you with a large 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. By connecting through Sahara, you are tapping into a potential community of early users; and these contributors are likely to become your product's real users 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 profit 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 in every link, including the invocation, combination, and cross-chain reuse after the model goes live, allowing value to be captured over a longer period.

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

This system brings about 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.

Because of this, I agree with Sahara's underlying belief: a truly healthy AI economic system cannot be merely about data plunder and the buyout of models, nor can it only benefit a select few. 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.

![AI × Web3: Who will build the chain for this era?])https://img-cdn.gateio.im/webp-social/moments-f100e78f0ea1234e20c00aebaa95457a.webp(

But 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 the project will face due to my investment position.

One of the 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: it supports deployment on any EVM-compatible chain and also provides standard API interfaces, allowing 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 poses 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 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 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 must 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 major trend of Web3 AI, but to truly unleash the value of its infrastructure, it still requires the implementation and improvement of more Web3 native AI products and technological solutions.

But don't forget, Sahara is currently "the only one of its kind".

In the blockchain infrastructure track originally designed for AI, although there are many imitators proposing conceptual frameworks, so far there is only Sah.